Skip to content
Python

Ten new apps for the Higgsfield surfaces still missing, and a workflow importer

neuregexed0ae4fcbf0e
diff --git a/apps/ai-influencer/app.json b/apps/ai-influencer/app.json
new file mode 100644
index 0000000..c05bed4
--- /dev/null
+++ b/apps/ai-influencer/app.json
@@ -0,0 +1,195 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "ai-influencer",
+ "name": "AI Influencer",
+ "section": "image",
+ "summary": "Makes a feed-ready photo of a made-up influencer in their niche, with natural skin (Z-Image Turbo).",
+ "higgsfield_feature": "Influencer IA",
+ "pack": "image",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "image_z_image_turbo_int8.json",
+ "comfyui": "0.42.0",
+ "notes": "Template app written by tools/template_apps.py: the workflow is apps/z-image's (built by tools/build_apps.py); the prompt below is fixed and \"Extra direction\" is appended to it."
+ },
+ "inputs": [
+ {
+ "id": "who",
+ "type": "text",
+ "label": "Who",
+ "default": "a woman in her late twenties from Mexico City with curly dark hair",
+ "required": true,
+ "help": "Describe a person who does not exist: age, origin, hair, style."
+ },
+ {
+ "id": "niche",
+ "type": "select",
+ "label": "Niche",
+ "default": "fitness influencer stretching on a rooftop at sunrise in workout clothes",
+ "options": [
+ {
+ "value": "fitness influencer stretching on a rooftop at sunrise in workout clothes",
+ "label": "Fitness"
+ },
+ {
+ "value": "travel influencer with a backpack at a viewpoint above a coastal town",
+ "label": "Travel"
+ },
+ {
+ "value": "fashion influencer posing on a city street in a tailored camel coat",
+ "label": "Fashion"
+ },
+ {
+ "value": "food influencer tasting a dish at a busy street food stall at night",
+ "label": "Food"
+ },
+ {
+ "value": "tech influencer at a desk with a laptop and a camera on a tripod, warm desk lamp",
+ "label": "Tech"
+ }
+ ]
+ },
+ {
+ "id": "details",
+ "type": "textarea",
+ "label": "Extra direction",
+ "default": "",
+ "help": "Added at the end of the app's prompt.",
+ "bind": [
+ {
+ "node": "57:27",
+ "field": "text",
+ "template": "A vertical phone photo for a social media feed of {who}, a {niche}. Natural skin texture with visible pores, candid moment, realistic photo, no text, no logos. {details}"
+ }
+ ]
+ },
+ {
+ "id": "width",
+ "type": "select",
+ "label": "Width",
+ "default": 768,
+ "options": [
+ {
+ "value": 768,
+ "label": "768"
+ },
+ {
+ "value": 1024,
+ "label": "1024"
+ },
+ {
+ "value": 1280,
+ "label": "1280"
+ },
+ {
+ "value": 1536,
+ "label": "1536"
+ }
+ ],
+ "bind": [
+ {
+ "node": "57:13",
+ "field": "width"
+ }
+ ]
+ },
+ {
+ "id": "height",
+ "type": "select",
+ "label": "Height",
+ "default": 1280,
+ "options": [
+ {
+ "value": 768,
+ "label": "768"
+ },
+ {
+ "value": 1024,
+ "label": "1024"
+ },
+ {
+ "value": 1280,
+ "label": "1280"
+ },
+ {
+ "value": 1536,
+ "label": "1536"
+ }
+ ],
+ "bind": [
+ {
+ "node": "57:13",
+ "field": "height"
+ }
+ ]
+ },
+ {
+ "id": "steps",
+ "type": "slider",
+ "label": "Steps",
+ "min": 4,
+ "max": 16,
+ "step": 1,
+ "default": 8,
+ "bind": [
+ {
+ "node": "57:3",
+ "field": "steps"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "57:3",
+ "field": "seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "9",
+ "type": "image",
+ "label": "Result"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/z_image_turbo",
+ "file": "split_files/diffusion_models/z_image_turbo_int8_convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 6.2,
+ "phase": "denoise",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/z_image_turbo",
+ "file": "split_files/text_encoders/qwen_3_4b_fp8_mixed.safetensors",
+ "dir": "text_encoders",
+ "gb": 5.63,
+ "phase": "encode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/z_image_turbo",
+ "file": "split_files/vae/ae.safetensors",
+ "dir": "vae",
+ "gb": 0.34,
+ "phase": "decode",
+ "license": "Apache-2.0"
+ }
+ ],
+ "credits": [
+ "Z-Image Turbo by Tongyi-MAI (Apache-2.0)"
+ ],
+ "test": {
+ "niche": "fitness influencer stretching on a rooftop at sunrise in workout clothes"
+ }
+}
diff --git a/apps/ai-influencer/workflow.api.json b/apps/ai-influencer/workflow.api.json
new file mode 100644
index 0000000..ca4e051
--- /dev/null
+++ b/apps/ai-influencer/workflow.api.json
@@ -0,0 +1,140 @@
+{
+ "9": {
+ "class_type": "SaveImage",
+ "inputs": {
+ "images": [
+ "57:8",
+ 0
+ ],
+ "filename_prefix": "freefield/ai-influencer"
+ },
+ "_meta": {
+ "title": "Save Image"
+ }
+ },
+ "57:3": {
+ "class_type": "KSampler",
+ "inputs": {
+ "model": [
+ "57:11",
+ 0
+ ],
+ "positive": [
+ "57:27",
+ 0
+ ],
+ "negative": [
+ "57:33",
+ 0
+ ],
+ "latent_image": [
+ "57:13",
+ 0
+ ],
+ "seed": 121725701057393,
+ "steps": 8,
+ "cfg": 1,
+ "sampler_name": "res_multistep",
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "KSampler"
+ }
+ },
+ "57:8": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "57:3",
+ 0
+ ],
+ "vae": [
+ "57:29",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "57:11": {
+ "class_type": "ModelSamplingAuraFlow",
+ "inputs": {
+ "model": [
+ "57:28",
+ 0
+ ],
+ "shift": 3
+ },
+ "_meta": {
+ "title": "ModelSamplingAuraFlow"
+ }
+ },
+ "57:13": {
+ "class_type": "EmptySD3LatentImage",
+ "inputs": {
+ "width": 1024,
+ "height": 1024,
+ "batch_size": 1
+ },
+ "_meta": {
+ "title": "EmptySD3LatentImage"
+ }
+ },
+ "57:27": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "57:30",
+ 0
+ ],
+ "text": "Dramatic black and white high fashion studio portrait, close-up bust shot, pale platinum blonde woman with sleek low ponytail, head tilted upward, eyes softly closed, wearing a fitted black turtleneck top. A large translucent pale white butterfly hovers gently right at her lips, delicate detailed wing veins visible. Hard rim light creates glowing bright white halo around her hair and face, deep inky pure black minimalist background, stark high contrast chiaroscuro lighting, film grain texture, moody ethereal atmosphere, monochrome, editorial fashion photography, shot on 35mm film, soft subtle skin texture, sharp focus on butterfly and facial profile, vertical composition, minimalist dark aesthetic, artistic surreal fashion"
+ },
+ "_meta": {
+ "title": "CLIP Text Encode (Prompt)"
+ }
+ },
+ "57:28": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "z_image_turbo_int8_convrot.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "57:29": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "ae.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "57:30": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "qwen_3_4b_fp8_mixed.safetensors",
+ "type": "lumina2",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "57:33": {
+ "class_type": "ConditioningZeroOut",
+ "inputs": {
+ "conditioning": [
+ "57:27",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Conditioning Zero Out"
+ }
+ }
+}
diff --git a/apps/bullet-time-scene/app.json b/apps/bullet-time-scene/app.json
new file mode 100644
index 0000000..8113a63
--- /dev/null
+++ b/apps/bullet-time-scene/app.json
@@ -0,0 +1,234 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "bullet-time-scene",
+ "name": "Bullet Time Scene",
+ "section": "video",
+ "summary": "Freezes the moment and circles the camera around the subject, in their own scene (Wan 2.1 + Bullet Time LoRA).",
+ "higgsfield_feature": "Bullet Time Scene",
+ "pack": "video-wan-effects",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "image_to_video_wan.json",
+ "comfyui": "0.42.0",
+ "notes": "Template app written by tools/template_apps.py: the workflow is apps/effects's (built by tools/build_apps.py) with Wan21_I2V_rotation_000003750.safetensors as the effect LoRA (node ff_effect); the prompt below is fixed and \"Extra direction\" is appended to it."
+ },
+ "inputs": [
+ {
+ "id": "image",
+ "type": "image",
+ "label": "Your photo",
+ "required": true,
+ "limits": {
+ "max_mb": 20
+ },
+ "bind": [
+ {
+ "node": "52",
+ "field": "image"
+ }
+ ]
+ },
+ {
+ "id": "subject",
+ "type": "text",
+ "label": "Who or what is in the photo",
+ "default": "a person mid-jump",
+ "required": true
+ },
+ {
+ "id": "details",
+ "type": "textarea",
+ "label": "Extra direction",
+ "default": "",
+ "help": "Added at the end of the app's prompt.",
+ "bind": [
+ {
+ "node": "6",
+ "field": "text",
+ "template": "rw360xz,人物定在空中,相机视角围绕人物向一个方向旋转。The video shows {subject}, frozen in mid-air while the camera circles around them in one direction; the whole scene is frozen too. {details}"
+ }
+ ]
+ },
+ {
+ "id": "strength",
+ "type": "slider",
+ "label": "Effect strength",
+ "min": 0.5,
+ "max": 1.4,
+ "step": 0.05,
+ "default": 1.0,
+ "bind": [
+ {
+ "node": "ff_effect",
+ "field": "strength_model"
+ }
+ ]
+ },
+ {
+ "id": "aspect",
+ "type": "select",
+ "label": "Format",
+ "default": "portrait",
+ "options": [
+ {
+ "value": "portrait",
+ "label": "9:16",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 480
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 832
+ }
+ ]
+ },
+ {
+ "value": "landscape",
+ "label": "16:9",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 832
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 480
+ }
+ ]
+ },
+ {
+ "value": "square",
+ "label": "1:1",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 624
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 624
+ }
+ ]
+ }
+ ],
+ "help": "Your photo is center-cropped to this shape."
+ },
+ {
+ "id": "frames",
+ "type": "select",
+ "label": "Length",
+ "default": 81,
+ "options": [
+ {
+ "value": 49,
+ "label": "3 s"
+ },
+ {
+ "value": 81,
+ "label": "5 s"
+ }
+ ],
+ "bind": [
+ {
+ "node": "50",
+ "field": "length"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "3",
+ "field": "seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "56",
+ "type": "video",
+ "label": "Result"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/diffusion_models/wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "dir": "diffusion_models",
+ "gb": 16.4,
+ "phase": "denoise",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors",
+ "dir": "text_encoders",
+ "gb": 6.74,
+ "phase": "encode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/clip_vision/clip_vision_h.safetensors",
+ "dir": "clip_vision",
+ "gb": 1.26,
+ "phase": "encode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/vae/wan_2.1_vae.safetensors",
+ "dir": "vae",
+ "gb": 0.25,
+ "phase": "decode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Kijai/WanVideo_comfy",
+ "file": "Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors",
+ "dir": "loras",
+ "gb": 0.74,
+ "phase": "denoise",
+ "license": "Apache-2.0",
+ "optional": true
+ },
+ {
+ "repo": "joyfox/Wan2.1-BulletTime-LoRA",
+ "file": "Wan21_I2V_rotation_000003750.safetensors",
+ "dir": "loras",
+ "gb": 0.36,
+ "phase": "denoise",
+ "license": "Apache-2.0",
+ "optional": true
+ }
+ ],
+ "credits": [
+ "Effect LoRAs by Remade-AI (Apache-2.0)",
+ "Bullet time, Money rain and Clones LoRAs by joyfox (Apache-2.0; the repos moved there from valiantcat and laozou)",
+ "Dolly zoom LoRA by ostris (Apache-2.0)",
+ "Steam LoRA by lvfengchun (Apache-2.0)",
+ "Wan 2.1 image-to-video 14B by Wan-AI, Alibaba (Apache-2.0)",
+ "Step-distill LoRA by lightx2v, packaged for ComfyUI by Kijai",
+ "Workflow: ComfyUI official template"
+ ],
+ "test": {
+ "image": "test/portrait.png",
+ "duration": 2,
+ "subject": "a parkour athlete in a black t-shirt leaping between two rooftops"
+ }
+}
diff --git a/apps/bullet-time-scene/workflow.api.json b/apps/bullet-time-scene/workflow.api.json
new file mode 100644
index 0000000..df7d058
--- /dev/null
+++ b/apps/bullet-time-scene/workflow.api.json
@@ -0,0 +1,240 @@
+{
+ "3": {
+ "class_type": "KSampler",
+ "inputs": {
+ "model": [
+ "54",
+ 0
+ ],
+ "positive": [
+ "50",
+ 0
+ ],
+ "negative": [
+ "50",
+ 1
+ ],
+ "latent_image": [
+ "50",
+ 2
+ ],
+ "seed": 987948718394761,
+ "steps": 20,
+ "cfg": 6,
+ "sampler_name": "uni_pc",
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "KSampler"
+ }
+ },
+ "6": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "38",
+ 0
+ ],
+ "text": "a cute anime girl with massive fennec ears and a big fluffy tail wearing a maid outfit turning around"
+ },
+ "_meta": {
+ "title": "CLIP Text Encode (Positive Prompt)"
+ }
+ },
+ "7": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "38",
+ 0
+ ],
+ "text": "色调艳丽,过曝,静态,细节模糊不清,字幕,风格,作品,画作,画面,静止,整体发灰,最差质量,低质量,JPEG压缩残留,丑陋的,残缺的,多余的手指,画得不好的手部,画得不好的脸部,畸形的,毁容的,形态畸形的肢体,手指融合,静止不动的画面,杂乱的背景,三条腿,背景人很多,倒着走"
+ },
+ "_meta": {
+ "title": "CLIP Text Encode (Negative Prompt)"
+ }
+ },
+ "8": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "3",
+ 0
+ ],
+ "vae": [
+ "39",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "37": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "38": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "umt5_xxl_fp8_e4m3fn_scaled.safetensors",
+ "type": "wan",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "39": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "wan_2.1_vae.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "49": {
+ "class_type": "CLIPVisionLoader",
+ "inputs": {
+ "clip_name": "clip_vision_h.safetensors"
+ },
+ "_meta": {
+ "title": "Load CLIP Vision"
+ }
+ },
+ "50": {
+ "class_type": "WanImageToVideo",
+ "inputs": {
+ "positive": [
+ "6",
+ 0
+ ],
+ "negative": [
+ "7",
+ 0
+ ],
+ "vae": [
+ "39",
+ 0
+ ],
+ "clip_vision_output": [
+ "51",
+ 0
+ ],
+ "start_image": [
+ "52",
+ 0
+ ],
+ "width": 480,
+ "height": 832,
+ "length": 81,
+ "batch_size": 1
+ },
+ "_meta": {
+ "title": "WanImageToVideo"
+ }
+ },
+ "51": {
+ "class_type": "CLIPVisionEncode",
+ "inputs": {
+ "clip_vision": [
+ "49",
+ 0
+ ],
+ "image": [
+ "52",
+ 0
+ ],
+ "crop": "none"
+ },
+ "_meta": {
+ "title": "CLIP Vision Encode"
+ }
+ },
+ "52": {
+ "class_type": "LoadImage",
+ "inputs": {
+ "image": "image_to_video_wan_start_image.png"
+ },
+ "_meta": {
+ "title": "Load Image"
+ }
+ },
+ "54": {
+ "class_type": "ModelSamplingSD3",
+ "inputs": {
+ "model": [
+ "ff_lora_fast",
+ 0
+ ],
+ "shift": 5.0
+ },
+ "_meta": {
+ "title": "ModelSamplingSD3"
+ }
+ },
+ "55": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "8",
+ 0
+ ],
+ "fps": 16
+ },
+ "_meta": {
+ "title": "Create Video"
+ }
+ },
+ "56": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "55",
+ 0
+ ],
+ "filename_prefix": "freefield/bullet-time-scene",
+ "format": "auto",
+ "format.codec": "auto"
+ },
+ "_meta": {
+ "title": "Save Video"
+ }
+ },
+ "ff_effect": {
+ "class_type": "LoraLoaderModelOnly",
+ "inputs": {
+ "model": [
+ "37",
+ 0
+ ],
+ "lora_name": "Wan21_I2V_rotation_000003750.safetensors",
+ "strength_model": 1.0
+ },
+ "_meta": {
+ "title": "Freefield LoRA"
+ }
+ },
+ "ff_lora_fast": {
+ "class_type": "LoraLoaderModelOnly",
+ "inputs": {
+ "model": [
+ "ff_effect",
+ 0
+ ],
+ "lora_name": "lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors",
+ "strength_model": 0.0
+ },
+ "_meta": {
+ "title": "Freefield LoRA"
+ }
+ }
+}
diff --git a/apps/bullet-time-splash/app.json b/apps/bullet-time-splash/app.json
new file mode 100644
index 0000000..59dd020
--- /dev/null
+++ b/apps/bullet-time-splash/app.json
@@ -0,0 +1,262 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "bullet-time-splash",
+ "name": "Bullet Time Splash",
+ "section": "video",
+ "summary": "Bursts frozen water around the subject, then circles the camera through the suspended drops.",
+ "higgsfield_feature": "Bullet Time Splash",
+ "pack": "video-wan-effects",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "image_to_video_wan.json",
+ "comfyui": "0.42.0",
+ "notes": "Template app written by tools/template_apps.py: the workflow is apps/effects's (built by tools/build_apps.py) with Wan21_I2V_rotation_000003750.safetensors as the effect LoRA (node ff_effect), after FLUX 3 Klein (apps/edit, merged in as edit:*) redraws the photo with a fixed instruction (node edit:4: \"Add a frozen splash of clear water bursting around this exact person, droplets and ribbons of water suspended in the air, same place and light. Keep the face, hair and outfit identical.\"); Wan starts from Klein's image; the prompt below is fixed and \"Extra direction\" is appended to it."
+ },
+ "inputs": [
+ {
+ "id": "image",
+ "type": "image",
+ "label": "Your photo",
+ "required": true,
+ "limits": {
+ "max_mb": 20
+ },
+ "bind": [
+ {
+ "node": "52",
+ "field": "image"
+ }
+ ]
+ },
+ {
+ "id": "subject",
+ "type": "text",
+ "label": "Who or what is in the photo",
+ "default": "a person",
+ "required": true
+ },
+ {
+ "id": "details",
+ "type": "textarea",
+ "label": "Extra direction",
+ "default": "",
+ "help": "Added at the end of the app's prompt.",
+ "bind": [
+ {
+ "node": "6",
+ "field": "text",
+ "template": "rw360xz,人物定在空中,相机视角围绕人物向一个方向旋转。The video shows {subject}, frozen in mid-air inside a burst of suspended water droplets while the camera circles around them in one direction. {details}"
+ }
+ ]
+ },
+ {
+ "id": "strength",
+ "type": "slider",
+ "label": "Effect strength",
+ "min": 0.5,
+ "max": 1.4,
+ "step": 0.05,
+ "default": 1.0,
+ "bind": [
+ {
+ "node": "ff_effect",
+ "field": "strength_model"
+ }
+ ]
+ },
+ {
+ "id": "aspect",
+ "type": "select",
+ "label": "Format",
+ "default": "portrait",
+ "options": [
+ {
+ "value": "portrait",
+ "label": "9:16",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 480
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 832
+ }
+ ]
+ },
+ {
+ "value": "landscape",
+ "label": "16:9",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 832
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 480
+ }
+ ]
+ },
+ {
+ "value": "square",
+ "label": "1:1",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 624
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 624
+ }
+ ]
+ }
+ ],
+ "help": "Your photo is center-cropped to this shape."
+ },
+ {
+ "id": "frames",
+ "type": "select",
+ "label": "Length",
+ "default": 81,
+ "options": [
+ {
+ "value": 49,
+ "label": "3 s"
+ },
+ {
+ "value": 81,
+ "label": "5 s"
+ }
+ ],
+ "bind": [
+ {
+ "node": "50",
+ "field": "length"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "3",
+ "field": "seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "56",
+ "type": "video",
+ "label": "Result"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/diffusion_models/wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "dir": "diffusion_models",
+ "gb": 16.4,
+ "phase": "denoise",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors",
+ "dir": "text_encoders",
+ "gb": 6.74,
+ "phase": "encode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/clip_vision/clip_vision_h.safetensors",
+ "dir": "clip_vision",
+ "gb": 1.26,
+ "phase": "encode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/vae/wan_2.1_vae.safetensors",
+ "dir": "vae",
+ "gb": 0.25,
+ "phase": "decode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Kijai/WanVideo_comfy",
+ "file": "Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors",
+ "dir": "loras",
+ "gb": 0.74,
+ "phase": "denoise",
+ "license": "Apache-2.0",
+ "optional": true
+ },
+ {
+ "repo": "joyfox/Wan2.1-BulletTime-LoRA",
+ "file": "Wan21_I2V_rotation_000003750.safetensors",
+ "dir": "loras",
+ "gb": 0.36,
+ "phase": "denoise",
+ "license": "Apache-2.0",
+ "optional": true
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "diffusion_models/flux3-image-klein-int8-convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 5.83,
+ "phase": "denoise",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "text_encoders/qwen3vl_4b_int8_convrot.safetensors",
+ "dir": "text_encoders",
+ "gb": 4.86,
+ "phase": "encode",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "vae/flux2-vae.safetensors",
+ "dir": "vae",
+ "gb": 0.34,
+ "phase": "decode",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ }
+ ],
+ "credits": [
+ "Effect LoRAs by Remade-AI (Apache-2.0)",
+ "Bullet time, Money rain and Clones LoRAs by joyfox (Apache-2.0; the repos moved there from valiantcat and laozou)",
+ "Dolly zoom LoRA by ostris (Apache-2.0)",
+ "Steam LoRA by lvfengchun (Apache-2.0)",
+ "Wan 2.1 image-to-video 14B by Wan-AI, Alibaba (Apache-2.0)",
+ "Step-distill LoRA by lightx2v, packaged for ComfyUI by Kijai",
+ "Workflow: ComfyUI official template",
+ "FLUX 3 Image Klein by Black Forest Labs (FLUX Kommunity License)"
+ ],
+ "test": {
+ "image": "test/portrait.png",
+ "duration": 2,
+ "subject": "a sprinter in a red and white running outfit"
+ }
+}
diff --git a/apps/bullet-time-splash/workflow.api.json b/apps/bullet-time-splash/workflow.api.json
new file mode 100644
index 0000000..e41a7de
--- /dev/null
+++ b/apps/bullet-time-splash/workflow.api.json
@@ -0,0 +1,367 @@
+{
+ "3": {
+ "class_type": "KSampler",
+ "inputs": {
+ "model": [
+ "54",
+ 0
+ ],
+ "positive": [
+ "50",
+ 0
+ ],
+ "negative": [
+ "50",
+ 1
+ ],
+ "latent_image": [
+ "50",
+ 2
+ ],
+ "seed": 987948718394761,
+ "steps": 20,
+ "cfg": 6,
+ "sampler_name": "uni_pc",
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "KSampler"
+ }
+ },
+ "6": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "38",
+ 0
+ ],
+ "text": "a cute anime girl with massive fennec ears and a big fluffy tail wearing a maid outfit turning around"
+ },
+ "_meta": {
+ "title": "CLIP Text Encode (Positive Prompt)"
+ }
+ },
+ "7": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "38",
+ 0
+ ],
+ "text": "色调艳丽,过曝,静态,细节模糊不清,字幕,风格,作品,画作,画面,静止,整体发灰,最差质量,低质量,JPEG压缩残留,丑陋的,残缺的,多余的手指,画得不好的手部,画得不好的脸部,畸形的,毁容的,形态畸形的肢体,手指融合,静止不动的画面,杂乱的背景,三条腿,背景人很多,倒着走"
+ },
+ "_meta": {
+ "title": "CLIP Text Encode (Negative Prompt)"
+ }
+ },
+ "8": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "3",
+ 0
+ ],
+ "vae": [
+ "39",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "37": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "38": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "umt5_xxl_fp8_e4m3fn_scaled.safetensors",
+ "type": "wan",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "39": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "wan_2.1_vae.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "49": {
+ "class_type": "CLIPVisionLoader",
+ "inputs": {
+ "clip_name": "clip_vision_h.safetensors"
+ },
+ "_meta": {
+ "title": "Load CLIP Vision"
+ }
+ },
+ "50": {
+ "class_type": "WanImageToVideo",
+ "inputs": {
+ "positive": [
+ "6",
+ 0
+ ],
+ "negative": [
+ "7",
+ 0
+ ],
+ "vae": [
+ "39",
+ 0
+ ],
+ "clip_vision_output": [
+ "51",
+ 0
+ ],
+ "start_image": [
+ "edit:8",
+ 0
+ ],
+ "width": 480,
+ "height": 832,
+ "length": 81,
+ "batch_size": 1
+ },
+ "_meta": {
+ "title": "WanImageToVideo"
+ }
+ },
+ "51": {
+ "class_type": "CLIPVisionEncode",
+ "inputs": {
+ "clip_vision": [
+ "49",
+ 0
+ ],
+ "image": [
+ "edit:8",
+ 0
+ ],
+ "crop": "none"
+ },
+ "_meta": {
+ "title": "CLIP Vision Encode"
+ }
+ },
+ "52": {
+ "class_type": "LoadImage",
+ "inputs": {
+ "image": "image_to_video_wan_start_image.png"
+ },
+ "_meta": {
+ "title": "Load Image"
+ }
+ },
+ "54": {
+ "class_type": "ModelSamplingSD3",
+ "inputs": {
+ "model": [
+ "ff_lora_fast",
+ 0
+ ],
+ "shift": 5.0
+ },
+ "_meta": {
+ "title": "ModelSamplingSD3"
+ }
+ },
+ "55": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "8",
+ 0
+ ],
+ "fps": 16
+ },
+ "_meta": {
+ "title": "Create Video"
+ }
+ },
+ "56": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "55",
+ 0
+ ],
+ "filename_prefix": "freefield/bullet-time-splash",
+ "format": "auto",
+ "format.codec": "auto"
+ },
+ "_meta": {
+ "title": "Save Video"
+ }
+ },
+ "ff_effect": {
+ "class_type": "LoraLoaderModelOnly",
+ "inputs": {
+ "model": [
+ "37",
+ 0
+ ],
+ "lora_name": "Wan21_I2V_rotation_000003750.safetensors",
+ "strength_model": 1.0
+ },
+ "_meta": {
+ "title": "Freefield LoRA"
+ }
+ },
+ "ff_lora_fast": {
+ "class_type": "LoraLoaderModelOnly",
+ "inputs": {
+ "model": [
+ "ff_effect",
+ 0
+ ],
+ "lora_name": "lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors",
+ "strength_model": 0.0
+ },
+ "_meta": {
+ "title": "Freefield LoRA"
+ }
+ },
+ "edit:1": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "flux3-image-klein-int8-convrot.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "edit:2": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "qwen3vl_4b_int8_convrot.safetensors",
+ "type": "flux3",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "edit:3": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "flux2-vae.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "edit:4": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "edit:2",
+ 0
+ ],
+ "text": "Add a frozen splash of clear water bursting around this exact person, droplets and ribbons of water suspended in the air, same place and light. Keep the face, hair and outfit identical."
+ },
+ "_meta": {
+ "title": "Positive Prompt"
+ }
+ },
+ "edit:7": {
+ "class_type": "KSampler",
+ "inputs": {
+ "model": [
+ "edit:1",
+ 0
+ ],
+ "positive": [
+ "edit:17",
+ 0
+ ],
+ "negative": [
+ "edit:17",
+ 1
+ ],
+ "latent_image": [
+ "edit:17",
+ 2
+ ],
+ "seed": 563228241303057,
+ "steps": 16,
+ "cfg": 1,
+ "sampler_name": "euler",
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "KSampler"
+ }
+ },
+ "edit:8": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "edit:7",
+ 0
+ ],
+ "vae": [
+ "edit:3",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "edit:17": {
+ "class_type": "ReferenceMultiLatent",
+ "inputs": {
+ "positive": [
+ "edit:4",
+ 0
+ ],
+ "negative": [
+ "edit:32",
+ 0
+ ],
+ "vae": [
+ "edit:3",
+ 0
+ ],
+ "images.image0": [
+ "52",
+ 0
+ ],
+ "resolution": "resize",
+ "resolution.megapixels": 1
+ },
+ "_meta": {
+ "title": "Set Reference Multi Latent"
+ }
+ },
+ "edit:32": {
+ "class_type": "ConditioningZeroOut",
+ "inputs": {
+ "conditioning": [
+ "edit:4",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Conditioning Zero Out"
+ }
+ }
+}
diff --git a/apps/bullet-time-white/app.json b/apps/bullet-time-white/app.json
new file mode 100644
index 0000000..350ed40
--- /dev/null
+++ b/apps/bullet-time-white/app.json
@@ -0,0 +1,262 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "bullet-time-white",
+ "name": "Bullet Time White",
+ "section": "video",
+ "summary": "Puts the subject mid-jump on a white studio backdrop, then freezes time and circles the camera around them.",
+ "higgsfield_feature": "Bullet Time White",
+ "pack": "video-wan-effects",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "image_to_video_wan.json",
+ "comfyui": "0.42.0",
+ "notes": "Template app written by tools/template_apps.py: the workflow is apps/effects's (built by tools/build_apps.py) with Wan21_I2V_rotation_000003750.safetensors as the effect LoRA (node ff_effect), after FLUX 3 Klein (apps/edit, merged in as edit:*) redraws the photo with a fixed instruction (node edit:4: \"Put this exact person on a seamless pure white studio background, mid-jump with arms spread, soft even studio light. Keep the face, hair and outfit identical.\"); Wan starts from Klein's image; the prompt below is fixed and \"Extra direction\" is appended to it."
+ },
+ "inputs": [
+ {
+ "id": "image",
+ "type": "image",
+ "label": "Your photo",
+ "required": true,
+ "limits": {
+ "max_mb": 20
+ },
+ "bind": [
+ {
+ "node": "52",
+ "field": "image"
+ }
+ ]
+ },
+ {
+ "id": "subject",
+ "type": "text",
+ "label": "Who or what is in the photo",
+ "default": "a person mid-jump",
+ "required": true
+ },
+ {
+ "id": "details",
+ "type": "textarea",
+ "label": "Extra direction",
+ "default": "",
+ "help": "Added at the end of the app's prompt.",
+ "bind": [
+ {
+ "node": "6",
+ "field": "text",
+ "template": "rw360xz,人物定在空中,相机视角围绕人物向一个方向旋转。The video shows {subject}, frozen in mid-air on a seamless white studio background while the camera circles around them in one direction. {details}"
+ }
+ ]
+ },
+ {
+ "id": "strength",
+ "type": "slider",
+ "label": "Effect strength",
+ "min": 0.5,
+ "max": 1.4,
+ "step": 0.05,
+ "default": 1.0,
+ "bind": [
+ {
+ "node": "ff_effect",
+ "field": "strength_model"
+ }
+ ]
+ },
+ {
+ "id": "aspect",
+ "type": "select",
+ "label": "Format",
+ "default": "portrait",
+ "options": [
+ {
+ "value": "portrait",
+ "label": "9:16",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 480
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 832
+ }
+ ]
+ },
+ {
+ "value": "landscape",
+ "label": "16:9",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 832
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 480
+ }
+ ]
+ },
+ {
+ "value": "square",
+ "label": "1:1",
+ "sets": [
+ {
+ "node": "50",
+ "field": "width",
+ "value": 624
+ },
+ {
+ "node": "50",
+ "field": "height",
+ "value": 624
+ }
+ ]
+ }
+ ],
+ "help": "Your photo is center-cropped to this shape."
+ },
+ {
+ "id": "frames",
+ "type": "select",
+ "label": "Length",
+ "default": 81,
+ "options": [
+ {
+ "value": 49,
+ "label": "3 s"
+ },
+ {
+ "value": 81,
+ "label": "5 s"
+ }
+ ],
+ "bind": [
+ {
+ "node": "50",
+ "field": "length"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "3",
+ "field": "seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "56",
+ "type": "video",
+ "label": "Result"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/diffusion_models/wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "dir": "diffusion_models",
+ "gb": 16.4,
+ "phase": "denoise",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors",
+ "dir": "text_encoders",
+ "gb": 6.74,
+ "phase": "encode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/clip_vision/clip_vision_h.safetensors",
+ "dir": "clip_vision",
+ "gb": 1.26,
+ "phase": "encode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
+ "file": "split_files/vae/wan_2.1_vae.safetensors",
+ "dir": "vae",
+ "gb": 0.25,
+ "phase": "decode",
+ "license": "Apache-2.0"
+ },
+ {
+ "repo": "Kijai/WanVideo_comfy",
+ "file": "Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors",
+ "dir": "loras",
+ "gb": 0.74,
+ "phase": "denoise",
+ "license": "Apache-2.0",
+ "optional": true
+ },
+ {
+ "repo": "joyfox/Wan2.1-BulletTime-LoRA",
+ "file": "Wan21_I2V_rotation_000003750.safetensors",
+ "dir": "loras",
+ "gb": 0.36,
+ "phase": "denoise",
+ "license": "Apache-2.0",
+ "optional": true
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "diffusion_models/flux3-image-klein-int8-convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 5.83,
+ "phase": "denoise",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "text_encoders/qwen3vl_4b_int8_convrot.safetensors",
+ "dir": "text_encoders",
+ "gb": 4.86,
+ "phase": "encode",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "vae/flux2-vae.safetensors",
+ "dir": "vae",
+ "gb": 0.34,
+ "phase": "decode",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ }
+ ],
+ "credits": [
+ "Effect LoRAs by Remade-AI (Apache-2.0)",
+ "Bullet time, Money rain and Clones LoRAs by joyfox (Apache-2.0; the repos moved there from valiantcat and laozou)",
+ "Dolly zoom LoRA by ostris (Apache-2.0)",
+ "Steam LoRA by lvfengchun (Apache-2.0)",
+ "Wan 2.1 image-to-video 14B by Wan-AI, Alibaba (Apache-2.0)",
+ "Step-distill LoRA by lightx2v, packaged for ComfyUI by Kijai",
+ "Workflow: ComfyUI official template",
+ "FLUX 3 Image Klein by Black Forest Labs (FLUX Kommunity License)"
+ ],
+ "test": {
+ "image": "test/portrait.png",
+ "duration": 2,
+ "subject": "a man in a red jacket, jeans and white sneakers"
+ }
+}
diff --git a/apps/bullet-time-white/workflow.api.json b/apps/bullet-time-white/workflow.api.json
new file mode 100644
index 0000000..4ef80b1
--- /dev/null
+++ b/apps/bullet-time-white/workflow.api.json
@@ -0,0 +1,367 @@
+{
+ "3": {
+ "class_type": "KSampler",
+ "inputs": {
+ "model": [
+ "54",
+ 0
+ ],
+ "positive": [
+ "50",
+ 0
+ ],
+ "negative": [
+ "50",
+ 1
+ ],
+ "latent_image": [
+ "50",
+ 2
+ ],
+ "seed": 987948718394761,
+ "steps": 20,
+ "cfg": 6,
+ "sampler_name": "uni_pc",
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "KSampler"
+ }
+ },
+ "6": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "38",
+ 0
+ ],
+ "text": "a cute anime girl with massive fennec ears and a big fluffy tail wearing a maid outfit turning around"
+ },
+ "_meta": {
+ "title": "CLIP Text Encode (Positive Prompt)"
+ }
+ },
+ "7": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "38",
+ 0
+ ],
+ "text": "色调艳丽,过曝,静态,细节模糊不清,字幕,风格,作品,画作,画面,静止,整体发灰,最差质量,低质量,JPEG压缩残留,丑陋的,残缺的,多余的手指,画得不好的手部,画得不好的脸部,畸形的,毁容的,形态畸形的肢体,手指融合,静止不动的画面,杂乱的背景,三条腿,背景人很多,倒着走"
+ },
+ "_meta": {
+ "title": "CLIP Text Encode (Negative Prompt)"
+ }
+ },
+ "8": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "3",
+ 0
+ ],
+ "vae": [
+ "39",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "37": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "38": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "umt5_xxl_fp8_e4m3fn_scaled.safetensors",
+ "type": "wan",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "39": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "wan_2.1_vae.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "49": {
+ "class_type": "CLIPVisionLoader",
+ "inputs": {
+ "clip_name": "clip_vision_h.safetensors"
+ },
+ "_meta": {
+ "title": "Load CLIP Vision"
+ }
+ },
+ "50": {
+ "class_type": "WanImageToVideo",
+ "inputs": {
+ "positive": [
+ "6",
+ 0
+ ],
+ "negative": [
+ "7",
+ 0
+ ],
+ "vae": [
+ "39",
+ 0
+ ],
+ "clip_vision_output": [
+ "51",
+ 0
+ ],
+ "start_image": [
+ "edit:8",
+ 0
+ ],
+ "width": 480,
+ "height": 832,
+ "length": 81,
+ "batch_size": 1
+ },
+ "_meta": {
+ "title": "WanImageToVideo"
+ }
+ },
+ "51": {
+ "class_type": "CLIPVisionEncode",
+ "inputs": {
+ "clip_vision": [
+ "49",
+ 0
+ ],
+ "image": [
+ "edit:8",
+ 0
+ ],
+ "crop": "none"
+ },
+ "_meta": {
+ "title": "CLIP Vision Encode"
+ }
+ },
+ "52": {
+ "class_type": "LoadImage",
+ "inputs": {
+ "image": "image_to_video_wan_start_image.png"
+ },
+ "_meta": {
+ "title": "Load Image"
+ }
+ },
+ "54": {
+ "class_type": "ModelSamplingSD3",
+ "inputs": {
+ "model": [
+ "ff_lora_fast",
+ 0
+ ],
+ "shift": 5.0
+ },
+ "_meta": {
+ "title": "ModelSamplingSD3"
+ }
+ },
+ "55": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "8",
+ 0
+ ],
+ "fps": 16
+ },
+ "_meta": {
+ "title": "Create Video"
+ }
+ },
+ "56": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "55",
+ 0
+ ],
+ "filename_prefix": "freefield/bullet-time-white",
+ "format": "auto",
+ "format.codec": "auto"
+ },
+ "_meta": {
+ "title": "Save Video"
+ }
+ },
+ "ff_effect": {
+ "class_type": "LoraLoaderModelOnly",
+ "inputs": {
+ "model": [
+ "37",
+ 0
+ ],
+ "lora_name": "Wan21_I2V_rotation_000003750.safetensors",
+ "strength_model": 1.0
+ },
+ "_meta": {
+ "title": "Freefield LoRA"
+ }
+ },
+ "ff_lora_fast": {
+ "class_type": "LoraLoaderModelOnly",
+ "inputs": {
+ "model": [
+ "ff_effect",
+ 0
+ ],
+ "lora_name": "lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors",
+ "strength_model": 0.0
+ },
+ "_meta": {
+ "title": "Freefield LoRA"
+ }
+ },
+ "edit:1": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "flux3-image-klein-int8-convrot.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "edit:2": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "qwen3vl_4b_int8_convrot.safetensors",
+ "type": "flux3",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "edit:3": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "flux2-vae.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "edit:4": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "edit:2",
+ 0
+ ],
+ "text": "Put this exact person on a seamless pure white studio background, mid-jump with arms spread, soft even studio light. Keep the face, hair and outfit identical."
+ },
+ "_meta": {
+ "title": "Positive Prompt"
+ }
+ },
+ "edit:7": {
+ "class_type": "KSampler",
+ "inputs": {
+ "model": [
+ "edit:1",
+ 0
+ ],
+ "positive": [
+ "edit:17",
+ 0
+ ],
+ "negative": [
+ "edit:17",
+ 1
+ ],
+ "latent_image": [
+ "edit:17",
+ 2
+ ],
+ "seed": 563228241303057,
+ "steps": 16,
+ "cfg": 1,
+ "sampler_name": "euler",
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "KSampler"
+ }
+ },
+ "edit:8": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "edit:7",
+ 0
+ ],
+ "vae": [
+ "edit:3",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "edit:17": {
+ "class_type": "ReferenceMultiLatent",
+ "inputs": {
+ "positive": [
+ "edit:4",
+ 0
+ ],
+ "negative": [
+ "edit:32",
+ 0
+ ],
+ "vae": [
+ "edit:3",
+ 0
+ ],
+ "images.image0": [
+ "52",
+ 0
+ ],
+ "resolution": "resize",
+ "resolution.megapixels": 1
+ },
+ "_meta": {
+ "title": "Set Reference Multi Latent"
+ }
+ },
+ "edit:32": {
+ "class_type": "ConditioningZeroOut",
+ "inputs": {
+ "conditioning": [
+ "edit:4",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Conditioning Zero Out"
+ }
+ }
+}
diff --git a/apps/depth-map/app.json b/apps/depth-map/app.json
new file mode 100644
index 0000000..d6f391e
--- /dev/null
+++ b/apps/depth-map/app.json
@@ -0,0 +1,102 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "depth-map",
+ "name": "Depth Map",
+ "section": "image",
+ "summary": "A depth map (or a normal map) of any photo, from MoGe-2's geometry estimate.",
+ "higgsfield_feature": "Depth Map",
+ "pack": "upscale",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "utility_moge_depth_estimation.json",
+ "comfyui": "0.42.0",
+ "notes": "Comfy's MoGe depth template: the colored depth render is saved, the grayscale and mask previews go; the map select picks MoGeRender's output."
+ },
+ "inputs": [
+ {
+ "id": "image",
+ "type": "image",
+ "label": "Photo",
+ "required": true,
+ "limits": {
+ "max_mb": 20
+ },
+ "bind": [
+ {
+ "node": "9",
+ "field": "image"
+ }
+ ]
+ },
+ {
+ "id": "map",
+ "type": "select",
+ "label": "Map",
+ "default": "depth_colored",
+ "options": [
+ {
+ "value": "depth_colored",
+ "label": "Depth, colored"
+ },
+ {
+ "value": "depth",
+ "label": "Depth, grayscale"
+ },
+ {
+ "value": "normal_opengl",
+ "label": "Normal map (OpenGL)"
+ },
+ {
+ "value": "normal_directx",
+ "label": "Normal map (DirectX)"
+ }
+ ],
+ "bind": [
+ {
+ "node": "49:23",
+ "field": "output"
+ }
+ ]
+ },
+ {
+ "id": "detail",
+ "type": "slider",
+ "label": "Detail",
+ "min": 0,
+ "max": 9,
+ "step": 1,
+ "default": 6,
+ "help": "Higher is sharper and slower.",
+ "bind": [
+ {
+ "node": "49:13",
+ "field": "resolution_level"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "17",
+ "type": "image",
+ "label": "Map"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/MoGe",
+ "file": "geometry_estimation/moge_2_vitl_normal_fp16.safetensors",
+ "dir": "geometry_estimation",
+ "gb": 0.66,
+ "phase": "all",
+ "license": "MIT"
+ }
+ ],
+ "credits": [
+ "MoGe-2 (Microsoft, MIT), ComfyUI port: Comfy-Org/MoGe"
+ ],
+ "test": {
+ "image": "test/portrait.png"
+ }
+}
diff --git a/apps/depth-map/workflow.api.json b/apps/depth-map/workflow.api.json
new file mode 100644
index 0000000..0ee1df3
--- /dev/null
+++ b/apps/depth-map/workflow.api.json
@@ -0,0 +1,145 @@
+{
+ "9": {
+ "class_type": "LoadImage",
+ "inputs": {
+ "image": "alien_world.png"
+ },
+ "_meta": {
+ "title": "Load Image"
+ }
+ },
+ "17": {
+ "class_type": "SaveImage",
+ "inputs": {
+ "images": [
+ "49:23",
+ 0
+ ],
+ "filename_prefix": "freefield/depth-map"
+ },
+ "_meta": {
+ "title": "Save Image"
+ }
+ },
+ "49:13": {
+ "class_type": "MoGeInference",
+ "inputs": {
+ "moge_model": [
+ "49:32",
+ 0
+ ],
+ "image": [
+ "49:53",
+ 0
+ ],
+ "resolution_level": 2,
+ "fov_x_degrees": 0,
+ "batch_size": 4,
+ "force_projection": true,
+ "apply_mask": true,
+ "refine_steps": 3
+ },
+ "_meta": {
+ "title": "Run MoGe Inference"
+ }
+ },
+ "49:23": {
+ "class_type": "MoGeRender",
+ "inputs": {
+ "moge_geometry": [
+ "49:13",
+ 0
+ ],
+ "output": "depth_colored"
+ },
+ "_meta": {
+ "title": "Render MoGe Geometry"
+ }
+ },
+ "49:32": {
+ "class_type": "LoadMoGeModel",
+ "inputs": {
+ "model_name": "moge_2_vitl_normal_fp16.safetensors"
+ },
+ "_meta": {
+ "title": "Load MoGe Model"
+ }
+ },
+ "49:36": {
+ "class_type": "ComfyMathExpression",
+ "inputs": {
+ "values.a": [
+ "49:37",
+ 0
+ ],
+ "expression": "a > 2048"
+ },
+ "_meta": {
+ "title": "Math Expression"
+ }
+ },
+ "49:37": {
+ "class_type": "GetImageSize",
+ "inputs": {
+ "image": [
+ "9",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Get Image Size"
+ }
+ },
+ "49:42": {
+ "class_type": "ComfySwitchNode",
+ "inputs": {
+ "on_false": [
+ "9",
+ 0
+ ],
+ "on_true": [
+ "49:70",
+ 0
+ ],
+ "switch": [
+ "49:36",
+ 2
+ ]
+ },
+ "_meta": {
+ "title": "If/Else Switch"
+ }
+ },
+ "49:53": {
+ "class_type": "ComfySwitchNode",
+ "inputs": {
+ "on_false": [
+ "9",
+ 0
+ ],
+ "on_true": [
+ "49:42",
+ 0
+ ],
+ "switch": true
+ },
+ "_meta": {
+ "title": "If/Else Switch"
+ }
+ },
+ "49:70": {
+ "class_type": "ResizeImageMaskNode",
+ "inputs": {
+ "input": [
+ "9",
+ 0
+ ],
+ "resize_type": "scale longer dimension",
+ "resize_type.longer_size": 2048,
+ "scale_method": "lanczos"
+ },
+ "_meta": {
+ "title": "Resize Image/Mask"
+ }
+ }
+}
diff --git a/apps/expand-image/app.json b/apps/expand-image/app.json
new file mode 100644
index 0000000..7dc2af5
--- /dev/null
+++ b/apps/expand-image/app.json
@@ -0,0 +1,204 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "expand-image",
+ "name": "Expand Image",
+ "section": "image",
+ "summary": "Grows a photo past its edges: FLUX 3 Image Klein draws only the new border, so the original stays as it was.",
+ "higgsfield_feature": "Expand Image",
+ "pack": "image-flux3",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "image_flux3_klein_distilled_image_edit.json",
+ "comfyui": "0.42.0",
+ "notes": "Template app written by tools/template_apps.py: the workflow is apps/edit's (built by tools/build_apps.py); the sampler starts from the photo scaled to 1 MP and padded by ImagePadForOutpaint (the select sets the padding), with a noise mask on the border only, and the unpadded photo stays Klein's reference; the prompt below is fixed and \"Extra direction\" is appended to it."
+ },
+ "inputs": [
+ {
+ "id": "image",
+ "type": "image",
+ "label": "Your photo",
+ "required": true,
+ "limits": {
+ "max_mb": 20
+ },
+ "bind": [
+ {
+ "node": "11",
+ "field": "image"
+ }
+ ]
+ },
+ {
+ "id": "direction",
+ "type": "select",
+ "label": "Expand",
+ "default": "out",
+ "options": [
+ {
+ "value": "out",
+ "label": "Zoom out",
+ "sets": [
+ {
+ "node": "op_pad",
+ "field": "left",
+ "value": 192
+ },
+ {
+ "node": "op_pad",
+ "field": "top",
+ "value": 192
+ },
+ {
+ "node": "op_pad",
+ "field": "right",
+ "value": 192
+ },
+ {
+ "node": "op_pad",
+ "field": "bottom",
+ "value": 192
+ }
+ ]
+ },
+ {
+ "value": "wide",
+ "label": "Wider",
+ "sets": [
+ {
+ "node": "op_pad",
+ "field": "left",
+ "value": 320
+ },
+ {
+ "node": "op_pad",
+ "field": "top",
+ "value": 0
+ },
+ {
+ "node": "op_pad",
+ "field": "right",
+ "value": 320
+ },
+ {
+ "node": "op_pad",
+ "field": "bottom",
+ "value": 0
+ }
+ ]
+ },
+ {
+ "value": "tall",
+ "label": "Taller",
+ "sets": [
+ {
+ "node": "op_pad",
+ "field": "left",
+ "value": 0
+ },
+ {
+ "node": "op_pad",
+ "field": "top",
+ "value": 320
+ },
+ {
+ "node": "op_pad",
+ "field": "right",
+ "value": 0
+ },
+ {
+ "node": "op_pad",
+ "field": "bottom",
+ "value": 320
+ }
+ ]
+ }
+ ]
+ },
+ {
+ "id": "details",
+ "type": "textarea",
+ "label": "Extra direction",
+ "default": "",
+ "help": "Added at the end of the app's prompt.",
+ "bind": [
+ {
+ "node": "4",
+ "field": "text",
+ "template": "Extend this photo outward into the gray border around it: continue the scene, its objects, light and perspective naturally past the original edges, as if the camera had pulled back. The original part of the photo stays exactly the same; no frame, no border, no text. {details}"
+ }
+ ]
+ },
+ {
+ "id": "steps",
+ "type": "slider",
+ "label": "Steps",
+ "min": 8,
+ "max": 32,
+ "step": 1,
+ "default": 16,
+ "bind": [
+ {
+ "node": "7",
+ "field": "steps"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "7",
+ "field": "seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "9",
+ "type": "image",
+ "label": "Result"
+ }
+ ],
+ "models": [
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "diffusion_models/flux3-image-klein-int8-convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 5.83,
+ "phase": "denoise",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "text_encoders/qwen3vl_4b_int8_convrot.safetensors",
+ "dir": "text_encoders",
+ "gb": 4.86,
+ "phase": "encode",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ },
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "vae/flux2-vae.safetensors",
+ "dir": "vae",
+ "gb": 0.34,
+ "phase": "decode",
+ "gated": true,
+ "license": "FLUX Kommunity License"
+ }
+ ],
+ "credits": [
+ "FLUX 3 Image Klein by Black Forest Labs (FLUX Kommunity License)"
+ ],
+ "test": {
+ "image": "test/portrait.png",
+ "direction": "out"
+ }
+}
diff --git a/apps/expand-image/workflow.api.json b/apps/expand-image/workflow.api.json
new file mode 100644
index 0000000..13a01f9
--- /dev/null
+++ b/apps/expand-image/workflow.api.json
@@ -0,0 +1,215 @@
+{
+ "1": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "flux3-image-klein-int8-convrot.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "2": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "qwen3vl_4b_int8_convrot.safetensors",
+ "type": "flux3",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "3": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "flux2-vae.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "4": {
+ "class_type": "CLIPTextEncode",
+ "inputs": {
+ "clip": [
+ "2",
+ 0
+ ],
+ "text": "Change the jacket to deep red leather, keep everything else the same."
+ },
+ "_meta": {
+ "title": "Positive Prompt"
+ }
+ },
+ "7": {
+ "class_type": "KSampler",
+ "inputs": {
+ "model": [
+ "1",
+ 0
+ ],
+ "positive": [
+ "17",
+ 0
+ ],
+ "negative": [
+ "17",
+ 1
+ ],
+ "latent_image": [
+ "op_latent",
+ 0
+ ],
+ "seed": 563228241303057,
+ "steps": 16,
+ "cfg": 1,
+ "sampler_name": "euler",
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "KSampler"
+ }
+ },
+ "8": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "7",
+ 0
+ ],
+ "vae": [
+ "3",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "9": {
+ "class_type": "SaveImage",
+ "inputs": {
+ "images": [
+ "8",
+ 0
+ ],
+ "filename_prefix": "freefield/expand-image"
+ },
+ "_meta": {
+ "title": "Save Image"
+ }
+ },
+ "11": {
+ "class_type": "LoadImage",
+ "inputs": {
+ "image": "woman_gold_earring_teal_top_portrait.png"
+ },
+ "_meta": {
+ "title": "Load Image"
+ }
+ },
+ "17": {
+ "class_type": "ReferenceMultiLatent",
+ "inputs": {
+ "positive": [
+ "4",
+ 0
+ ],
+ "negative": [
+ "32",
+ 0
+ ],
+ "vae": [
+ "3",
+ 0
+ ],
+ "images.image0": [
+ "11",
+ 0
+ ],
+ "resolution": "resize",
+ "resolution.megapixels": 1
+ },
+ "_meta": {
+ "title": "Set Reference Multi Latent"
+ }
+ },
+ "32": {
+ "class_type": "ConditioningZeroOut",
+ "inputs": {
+ "conditioning": [
+ "4",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Conditioning Zero Out"
+ }
+ },
+ "op_scale": {
+ "class_type": "ImageScaleToTotalPixels",
+ "inputs": {
+ "image": [
+ "11",
+ 0
+ ],
+ "upscale_method": "lanczos",
+ "megapixels": 1.0,
+ "resolution_steps": 16
+ },
+ "_meta": {
+ "title": "Photo at 1 MP"
+ }
+ },
+ "op_pad": {
+ "class_type": "ImagePadForOutpaint",
+ "inputs": {
+ "image": [
+ "op_scale",
+ 0
+ ],
+ "left": 192,
+ "top": 192,
+ "right": 192,
+ "bottom": 192,
+ "feathering": 24
+ },
+ "_meta": {
+ "title": "Border to draw"
+ }
+ },
+ "op_encode": {
+ "class_type": "VAEEncode",
+ "inputs": {
+ "pixels": [
+ "op_pad",
+ 0
+ ],
+ "vae": [
+ "3",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Encode the padded photo"
+ }
+ },
+ "op_latent": {
+ "class_type": "SetLatentNoiseMask",
+ "inputs": {
+ "samples": [
+ "op_encode",
+ 0
+ ],
+ "mask": [
+ "op_pad",
+ 1
+ ]
+ },
+ "_meta": {
+ "title": "Only the border is drawn"
+ }
+ }
+}
diff --git a/apps/extend-video/app.json b/apps/extend-video/app.json
new file mode 100644
index 0000000..382870a
--- /dev/null
+++ b/apps/extend-video/app.json
@@ -0,0 +1,139 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "extend-video",
+ "name": "Extend Video",
+ "section": "video",
+ "summary": "Continues a clip from its last frame with MiniMax H3, sound included, and joins the new part after it.",
+ "higgsfield_feature": "Extend Video",
+ "pack": "video-h3",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "video_minimax_h3_i2v.json",
+ "comfyui": "0.42.0",
+ "notes": "Comfy's video_minimax_h3_i2v template with the int8 text encoder in place of the NVFP4 one (it needs a Blackwell GPU) and the plain 20-step branch of its turbo switch wired in: lightx2v's turbo LoRA is left out, its model card states no license (tools/build_apps.py). The clip's last frame (ImageFromBatch -1) starts the new part; ImageBatch joins the frames (the new ones scaled to the clip's size), AudioConcat the sound, CreateVideo writes them at the clip's frame rate."
+ },
+ "inputs": [
+ {
+ "id": "clip",
+ "type": "video",
+ "label": "Your clip",
+ "required": true,
+ "limits": {
+ "max_mb": 60
+ },
+ "help": "Up to about 15 s; 24 or 25 fps clips join best.",
+ "bind": [
+ {
+ "node": "ff_clip",
+ "field": "file"
+ }
+ ]
+ },
+ {
+ "id": "prompt",
+ "type": "textarea",
+ "label": "What happens next",
+ "required": true,
+ "default": "The scene continues naturally: same people, place, light and camera style.",
+ "bind": [
+ {
+ "node": "105:104",
+ "field": "prompt"
+ }
+ ]
+ },
+ {
+ "id": "duration",
+ "type": "slider",
+ "label": "Seconds to add",
+ "min": 2,
+ "max": 10,
+ "step": 1,
+ "default": 3,
+ "bind": [
+ {
+ "node": "105:111",
+ "field": "value"
+ }
+ ]
+ },
+ {
+ "id": "steps",
+ "type": "slider",
+ "label": "Steps",
+ "min": 8,
+ "max": 30,
+ "step": 1,
+ "default": 20,
+ "help": "Fewer steps run faster and look rougher.",
+ "bind": [
+ {
+ "node": "105:124",
+ "field": "value"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "105:15",
+ "field": "noise_seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "92",
+ "type": "video",
+ "label": "Extended clip"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 20.97,
+ "phase": "denoise",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "text_encoders/qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "dir": "text_encoders",
+ "gb": 27.14,
+ "phase": "encode",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "vae/minimax_h3_video_vae_int8_convrot.safetensors",
+ "dir": "vae",
+ "gb": 2.81,
+ "phase": "decode",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "vae/minimax_h3_audio_vae_fp32.safetensors",
+ "dir": "vae",
+ "gb": 0.61,
+ "phase": "decode",
+ "license": "MiniMax H3 Community License"
+ }
+ ],
+ "credits": [
+ "MiniMax H3 (MiniMax H3 Community License)"
+ ],
+ "test": {
+ "clip": "test/clip.mp4",
+ "duration": 2
+ }
+}
diff --git a/apps/extend-video/workflow.api.json b/apps/extend-video/workflow.api.json
new file mode 100644
index 0000000..bf12660
--- /dev/null
+++ b/apps/extend-video/workflow.api.json
@@ -0,0 +1,348 @@
+{
+ "92": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "105:91",
+ 0
+ ],
+ "filename_prefix": "freefield/extend-video",
+ "format": "auto",
+ "format.codec": "auto"
+ },
+ "_meta": {
+ "title": "Save Video"
+ }
+ },
+ "105:6": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "105:9": {
+ "class_type": "BasicScheduler",
+ "inputs": {
+ "model": [
+ "105:6",
+ 0
+ ],
+ "steps": [
+ "105:124",
+ 0
+ ],
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "BasicScheduler"
+ }
+ },
+ "105:10": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "105:14",
+ 0
+ ],
+ "vae": [
+ "105:11",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "105:11": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "minimax_h3_video_vae_int8_convrot.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "105:13": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "type": "minimax",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "105:14": {
+ "class_type": "SamplerCustomAdvanced",
+ "inputs": {
+ "noise": [
+ "105:15",
+ 0
+ ],
+ "guider": [
+ "105:16",
+ 0
+ ],
+ "sampler": [
+ "105:17",
+ 0
+ ],
+ "sigmas": [
+ "105:9",
+ 0
+ ],
+ "latent_image": [
+ "105:104",
+ 1
+ ]
+ },
+ "_meta": {
+ "title": "SamplerCustomAdvanced"
+ }
+ },
+ "105:15": {
+ "class_type": "RandomNoise",
+ "inputs": {
+ "noise_seed": 1
+ },
+ "_meta": {
+ "title": "RandomNoise"
+ }
+ },
+ "105:16": {
+ "class_type": "BasicGuider",
+ "inputs": {
+ "model": [
+ "105:6",
+ 0
+ ],
+ "conditioning": [
+ "105:104",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Basic Guider"
+ }
+ },
+ "105:17": {
+ "class_type": "KSamplerSelect",
+ "inputs": {
+ "sampler_name": "res_multistep"
+ },
+ "_meta": {
+ "title": "KSamplerSelect"
+ }
+ },
+ "105:23": {
+ "class_type": "VAEDecodeAudio",
+ "inputs": {
+ "samples": [
+ "105:14",
+ 0
+ ],
+ "vae": [
+ "105:24",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode Audio"
+ }
+ },
+ "105:24": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "minimax_h3_audio_vae_fp32.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "105:91": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "ff_frames",
+ 0
+ ],
+ "audio": [
+ "ff_sound",
+ 0
+ ],
+ "fps": [
+ "ff_parts",
+ 2
+ ],
+ "bit_depth": 8
+ },
+ "_meta": {
+ "title": "Create Video"
+ }
+ },
+ "105:104": {
+ "class_type": "MiniMaxH3ImageToVideo",
+ "inputs": {
+ "clip": [
+ "105:13",
+ 0
+ ],
+ "vae": [
+ "105:11",
+ 0
+ ],
+ "first_frame": [
+ "119",
+ 0
+ ],
+ "width": [
+ "120",
+ 0
+ ],
+ "height": [
+ "120",
+ 1
+ ],
+ "length": [
+ "105:107",
+ 1
+ ],
+ "prompt": "Vaporwave title sequence look: pink and blue gradient palette, VHS tracking artifacts, Greek statue motifs, chrome palm trees, RGB chromatic aberration, lo-fi retro atmosphere, mood languid and nostalgic.\n\nTimeline:\n[0s-1s] VHS static opens the frame, the title \"COMFYUI\" appears with RGB split and a slight horizontal jitter.\n[1s-2.5s] Hard cut, a Greek plaster bust close-up, pink-purple gradient sky, a pixelated sun.\n[2.5s-4s] Clean \"STARRING\" credits appear, \"LATENT\" and \"CONTROLNET\" each shown exactly once.\n[4s-5s] Final card \"DIRECTED BY COMFYUI\" holds, one VHS tracking glitch settling into stability.\n\nHard cuts only, transitions landing with tape jumps, no push-ins, no dissolves.\n\nAudio: lo-fi vaporwave score, slow drum machine with soft bass, VHS tape-noise sample joins at 2.5s, melody fading for the last 1s.\n\nAll text must be clearly legible, do not misspell English, no Chinese characters, do not repeat names or job titles, no soft dissolves, no subtitle bars."
+ },
+ "_meta": {
+ "title": "MiniMax H3 Image to Video"
+ }
+ },
+ "105:107": {
+ "class_type": "ComfyMathExpression",
+ "inputs": {
+ "values.a": [
+ "105:111",
+ 0
+ ],
+ "expression": "max(5, round(a * 24)) + (5 - (max(5, round(a * 24)) % 17)) % 17"
+ },
+ "_meta": {
+ "title": "Math Expression"
+ }
+ },
+ "105:111": {
+ "class_type": "PrimitiveFloat",
+ "inputs": {
+ "value": 2
+ },
+ "_meta": {
+ "title": "Float (duration)"
+ }
+ },
+ "105:124": {
+ "class_type": "PrimitiveInt",
+ "inputs": {
+ "value": 20
+ },
+ "_meta": {
+ "title": "Int"
+ }
+ },
+ "119": {
+ "class_type": "ImageScaleToTotalPixels",
+ "inputs": {
+ "upscale_method": "lanczos",
+ "megapixels": 0.6,
+ "resolution_steps": 32,
+ "image": [
+ "ff_last",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Scale Image to Total Pixels"
+ }
+ },
+ "120": {
+ "class_type": "GetImageSize",
+ "inputs": {
+ "image": [
+ "119",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Get Image Size"
+ }
+ },
+ "ff_clip": {
+ "class_type": "LoadVideo",
+ "inputs": {
+ "file": "clip.mp4"
+ },
+ "_meta": {
+ "title": "Clip to extend"
+ }
+ },
+ "ff_parts": {
+ "class_type": "GetVideoComponents",
+ "inputs": {
+ "video": [
+ "ff_clip",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Frames, sound, fps"
+ }
+ },
+ "ff_last": {
+ "class_type": "ImageFromBatch",
+ "inputs": {
+ "image": [
+ "ff_parts",
+ 0
+ ],
+ "batch_index": -1,
+ "length": 1
+ },
+ "_meta": {
+ "title": "Last frame"
+ }
+ },
+ "ff_frames": {
+ "class_type": "ImageBatch",
+ "inputs": {
+ "image1": [
+ "ff_parts",
+ 0
+ ],
+ "image2": [
+ "105:10",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Clip + continuation"
+ }
+ },
+ "ff_sound": {
+ "class_type": "AudioConcat",
+ "inputs": {
+ "audio1": [
+ "ff_parts",
+ 1
+ ],
+ "audio2": [
+ "105:23",
+ 0
+ ],
+ "direction": "after"
+ },
+ "_meta": {
+ "title": "Sound + new sound"
+ }
+ }
+}
diff --git a/apps/h3-photo/app.json b/apps/h3-photo/app.json
new file mode 100644
index 0000000..c7b7803
--- /dev/null
+++ b/apps/h3-photo/app.json
@@ -0,0 +1,140 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "h3-photo",
+ "name": "H3 Photo to Video",
+ "section": "video",
+ "summary": "Animates a photo into a clip with its own sound on MiniMax H3; the clip keeps the photo's shape.",
+ "higgsfield_feature": "Create Video",
+ "pack": "video-h3",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "video_minimax_h3_i2v.json",
+ "comfyui": "0.42.0",
+ "notes": "Comfy's video_minimax_h3_i2v template with the int8 text encoder in place of the NVFP4 one (it needs a Blackwell GPU) and the plain 20-step branch of its turbo switch wired in: lightx2v's turbo LoRA is left out, its model card states no license (tools/build_apps.py). The photo, scaled to 0.6 MP in steps of 32, sets the clip's size."
+ },
+ "inputs": [
+ {
+ "id": "image",
+ "type": "image",
+ "label": "Photo",
+ "required": true,
+ "limits": {
+ "max_mb": 20
+ },
+ "bind": [
+ {
+ "node": "114",
+ "field": "image"
+ }
+ ]
+ },
+ {
+ "id": "prompt",
+ "type": "textarea",
+ "label": "What happens",
+ "required": true,
+ "default": "The person smiles, waves at the camera and says: \"Hey, good to see you.\"",
+ "help": "Put spoken words in quotes; describe sounds and music too.",
+ "bind": [
+ {
+ "node": "105:104",
+ "field": "prompt"
+ }
+ ]
+ },
+ {
+ "id": "duration",
+ "type": "slider",
+ "label": "Duration (s)",
+ "min": 2,
+ "max": 10,
+ "step": 1,
+ "default": 4,
+ "bind": [
+ {
+ "node": "105:111",
+ "field": "value"
+ }
+ ]
+ },
+ {
+ "id": "steps",
+ "type": "slider",
+ "label": "Steps",
+ "min": 8,
+ "max": 30,
+ "step": 1,
+ "default": 20,
+ "help": "Fewer steps run faster and look rougher.",
+ "bind": [
+ {
+ "node": "105:124",
+ "field": "value"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "105:15",
+ "field": "noise_seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "92",
+ "type": "video",
+ "label": "Video"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 20.97,
+ "phase": "denoise",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "text_encoders/qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "dir": "text_encoders",
+ "gb": 27.14,
+ "phase": "encode",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "vae/minimax_h3_video_vae_int8_convrot.safetensors",
+ "dir": "vae",
+ "gb": 2.81,
+ "phase": "decode",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "vae/minimax_h3_audio_vae_fp32.safetensors",
+ "dir": "vae",
+ "gb": 0.61,
+ "phase": "decode",
+ "license": "MiniMax H3 Community License"
+ }
+ ],
+ "credits": [
+ "MiniMax H3 (MiniMax H3 Community License)"
+ ],
+ "test": {
+ "image": "test/portrait.png",
+ "prompt": "she smiles",
+ "duration": 2
+ }
+}
diff --git a/apps/h3-photo/workflow.api.json b/apps/h3-photo/workflow.api.json
new file mode 100644
index 0000000..ababaa8
--- /dev/null
+++ b/apps/h3-photo/workflow.api.json
@@ -0,0 +1,286 @@
+{
+ "92": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "105:91",
+ 0
+ ],
+ "filename_prefix": "freefield/h3-photo",
+ "format": "auto",
+ "format.codec": "auto"
+ },
+ "_meta": {
+ "title": "Save Video"
+ }
+ },
+ "105:6": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "105:9": {
+ "class_type": "BasicScheduler",
+ "inputs": {
+ "model": [
+ "105:6",
+ 0
+ ],
+ "steps": [
+ "105:124",
+ 0
+ ],
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "BasicScheduler"
+ }
+ },
+ "105:10": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "105:14",
+ 0
+ ],
+ "vae": [
+ "105:11",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "105:11": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "minimax_h3_video_vae_int8_convrot.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "105:13": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "type": "minimax",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "105:14": {
+ "class_type": "SamplerCustomAdvanced",
+ "inputs": {
+ "noise": [
+ "105:15",
+ 0
+ ],
+ "guider": [
+ "105:16",
+ 0
+ ],
+ "sampler": [
+ "105:17",
+ 0
+ ],
+ "sigmas": [
+ "105:9",
+ 0
+ ],
+ "latent_image": [
+ "105:104",
+ 1
+ ]
+ },
+ "_meta": {
+ "title": "SamplerCustomAdvanced"
+ }
+ },
+ "105:15": {
+ "class_type": "RandomNoise",
+ "inputs": {
+ "noise_seed": 1
+ },
+ "_meta": {
+ "title": "RandomNoise"
+ }
+ },
+ "105:16": {
+ "class_type": "BasicGuider",
+ "inputs": {
+ "model": [
+ "105:6",
+ 0
+ ],
+ "conditioning": [
+ "105:104",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Basic Guider"
+ }
+ },
+ "105:17": {
+ "class_type": "KSamplerSelect",
+ "inputs": {
+ "sampler_name": "res_multistep"
+ },
+ "_meta": {
+ "title": "KSamplerSelect"
+ }
+ },
+ "105:23": {
+ "class_type": "VAEDecodeAudio",
+ "inputs": {
+ "samples": [
+ "105:14",
+ 0
+ ],
+ "vae": [
+ "105:24",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode Audio"
+ }
+ },
+ "105:24": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "minimax_h3_audio_vae_fp32.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "105:91": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "105:10",
+ 0
+ ],
+ "audio": [
+ "105:23",
+ 0
+ ],
+ "fps": 24,
+ "bit_depth": 8
+ },
+ "_meta": {
+ "title": "Create Video"
+ }
+ },
+ "105:104": {
+ "class_type": "MiniMaxH3ImageToVideo",
+ "inputs": {
+ "clip": [
+ "105:13",
+ 0
+ ],
+ "vae": [
+ "105:11",
+ 0
+ ],
+ "first_frame": [
+ "119",
+ 0
+ ],
+ "width": [
+ "120",
+ 0
+ ],
+ "height": [
+ "120",
+ 1
+ ],
+ "length": [
+ "105:107",
+ 1
+ ],
+ "prompt": "Vaporwave title sequence look: pink and blue gradient palette, VHS tracking artifacts, Greek statue motifs, chrome palm trees, RGB chromatic aberration, lo-fi retro atmosphere, mood languid and nostalgic.\n\nTimeline:\n[0s-1s] VHS static opens the frame, the title \"COMFYUI\" appears with RGB split and a slight horizontal jitter.\n[1s-2.5s] Hard cut, a Greek plaster bust close-up, pink-purple gradient sky, a pixelated sun.\n[2.5s-4s] Clean \"STARRING\" credits appear, \"LATENT\" and \"CONTROLNET\" each shown exactly once.\n[4s-5s] Final card \"DIRECTED BY COMFYUI\" holds, one VHS tracking glitch settling into stability.\n\nHard cuts only, transitions landing with tape jumps, no push-ins, no dissolves.\n\nAudio: lo-fi vaporwave score, slow drum machine with soft bass, VHS tape-noise sample joins at 2.5s, melody fading for the last 1s.\n\nAll text must be clearly legible, do not misspell English, no Chinese characters, do not repeat names or job titles, no soft dissolves, no subtitle bars."
+ },
+ "_meta": {
+ "title": "MiniMax H3 Image to Video"
+ }
+ },
+ "105:107": {
+ "class_type": "ComfyMathExpression",
+ "inputs": {
+ "values.a": [
+ "105:111",
+ 0
+ ],
+ "expression": "max(5, round(a * 24)) + (5 - (max(5, round(a * 24)) % 17)) % 17"
+ },
+ "_meta": {
+ "title": "Math Expression"
+ }
+ },
+ "105:111": {
+ "class_type": "PrimitiveFloat",
+ "inputs": {
+ "value": 2
+ },
+ "_meta": {
+ "title": "Float (duration)"
+ }
+ },
+ "105:124": {
+ "class_type": "PrimitiveInt",
+ "inputs": {
+ "value": 20
+ },
+ "_meta": {
+ "title": "Int"
+ }
+ },
+ "114": {
+ "class_type": "LoadImage",
+ "inputs": {
+ "image": "transparent_rgb_gaming_mouse.png"
+ },
+ "_meta": {
+ "title": "Load Image"
+ }
+ },
+ "119": {
+ "class_type": "ImageScaleToTotalPixels",
+ "inputs": {
+ "upscale_method": "lanczos",
+ "megapixels": 0.6,
+ "resolution_steps": 32,
+ "image": [
+ "114",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Scale Image to Total Pixels"
+ }
+ },
+ "120": {
+ "class_type": "GetImageSize",
+ "inputs": {
+ "image": [
+ "119",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Get Image Size"
+ }
+ }
+}
diff --git a/apps/h3-video/app.json b/apps/h3-video/app.json
new file mode 100644
index 0000000..70062a3
--- /dev/null
+++ b/apps/h3-video/app.json
@@ -0,0 +1,154 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "h3-video",
+ "name": "H3 Video",
+ "section": "video",
+ "summary": "A clip with its own sound from a prompt, on MiniMax H3: speech in quotes, sound effects and music come out of the same model.",
+ "higgsfield_feature": "Create Video",
+ "pack": "video-h3",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "video_minimax_h3_t2v.json",
+ "comfyui": "0.42.0",
+ "notes": "Comfy's video_minimax_h3_t2v template with the int8 text encoder in place of the NVFP4 one (it needs a Blackwell GPU) and the plain 20-step branch of its turbo switch wired in: lightx2v's turbo LoRA is left out, its model card states no license (tools/build_apps.py)."
+ },
+ "inputs": [
+ {
+ "id": "prompt",
+ "type": "textarea",
+ "label": "Prompt",
+ "required": true,
+ "default": "A street cook in Mexico City flips tacos on a hot griddle, steam and sizzling sound, he looks up and says: \"Two more coming right up!\"",
+ "help": "Put spoken words in quotes; describe sounds and music too.",
+ "bind": [
+ {
+ "node": "140:131",
+ "field": "prompt"
+ }
+ ]
+ },
+ {
+ "id": "aspect",
+ "type": "select",
+ "label": "Format",
+ "default": "16:9 (Widescreen)",
+ "options": [
+ {
+ "value": "16:9 (Widescreen)",
+ "label": "16:9"
+ },
+ {
+ "value": "9:16 (Portrait Widescreen)",
+ "label": "9:16"
+ },
+ {
+ "value": "1:1 (Square)",
+ "label": "1:1"
+ },
+ {
+ "value": "4:3 (Standard)",
+ "label": "4:3"
+ }
+ ],
+ "bind": [
+ {
+ "node": "115",
+ "field": "aspect_ratio"
+ }
+ ]
+ },
+ {
+ "id": "duration",
+ "type": "slider",
+ "label": "Duration (s)",
+ "min": 2,
+ "max": 10,
+ "step": 1,
+ "default": 4,
+ "bind": [
+ {
+ "node": "140:133",
+ "field": "value"
+ }
+ ]
+ },
+ {
+ "id": "steps",
+ "type": "slider",
+ "label": "Steps",
+ "min": 8,
+ "max": 30,
+ "step": 1,
+ "default": 20,
+ "help": "Fewer steps run faster and look rougher.",
+ "bind": [
+ {
+ "node": "140:137",
+ "field": "value"
+ }
+ ]
+ },
+ {
+ "id": "seed",
+ "type": "seed",
+ "label": "Seed",
+ "default": -1,
+ "help": "-1 picks a random seed",
+ "bind": [
+ {
+ "node": "140:129",
+ "field": "noise_seed"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "92",
+ "type": "video",
+ "label": "Video"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 20.97,
+ "phase": "denoise",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "text_encoders/qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "dir": "text_encoders",
+ "gb": 27.14,
+ "phase": "encode",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "vae/minimax_h3_video_vae_int8_convrot.safetensors",
+ "dir": "vae",
+ "gb": 2.81,
+ "phase": "decode",
+ "license": "MiniMax H3 Community License"
+ },
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "vae/minimax_h3_audio_vae_fp32.safetensors",
+ "dir": "vae",
+ "gb": 0.61,
+ "phase": "decode",
+ "license": "MiniMax H3 Community License"
+ }
+ ],
+ "credits": [
+ "MiniMax H3 (MiniMax H3 Community License)"
+ ],
+ "test": {
+ "prompt": "a cat yawns on a sunny windowsill",
+ "duration": 2
+ }
+}
diff --git a/apps/h3-video/workflow.api.json b/apps/h3-video/workflow.api.json
new file mode 100644
index 0000000..74ab894
--- /dev/null
+++ b/apps/h3-video/workflow.api.json
@@ -0,0 +1,257 @@
+{
+ "92": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "140:130",
+ 0
+ ],
+ "filename_prefix": "freefield/h3-video",
+ "format": "auto",
+ "format.codec": "auto"
+ },
+ "_meta": {
+ "title": "Save Video"
+ }
+ },
+ "115": {
+ "class_type": "ResolutionSelector",
+ "inputs": {
+ "aspect_ratio": "16:9 (Widescreen)",
+ "megapixels": 0.4,
+ "multiple": 32
+ },
+ "_meta": {
+ "title": "Resolution Selector"
+ }
+ },
+ "140:119": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "minimax_h3_video_vae_int8_convrot.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "140:120": {
+ "class_type": "VAELoader",
+ "inputs": {
+ "vae_name": "minimax_h3_audio_vae_fp32.safetensors"
+ },
+ "_meta": {
+ "title": "Load VAE"
+ }
+ },
+ "140:121": {
+ "class_type": "VAEDecodeAudio",
+ "inputs": {
+ "samples": [
+ "140:125",
+ 0
+ ],
+ "vae": [
+ "140:120",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode Audio"
+ }
+ },
+ "140:122": {
+ "class_type": "VAEDecode",
+ "inputs": {
+ "samples": [
+ "140:125",
+ 0
+ ],
+ "vae": [
+ "140:119",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "VAE Decode"
+ }
+ },
+ "140:123": {
+ "class_type": "KSamplerSelect",
+ "inputs": {
+ "sampler_name": "res_multistep"
+ },
+ "_meta": {
+ "title": "KSamplerSelect"
+ }
+ },
+ "140:124": {
+ "class_type": "BasicScheduler",
+ "inputs": {
+ "model": [
+ "140:127",
+ 0
+ ],
+ "steps": [
+ "140:137",
+ 0
+ ],
+ "scheduler": "simple",
+ "denoise": 1
+ },
+ "_meta": {
+ "title": "BasicScheduler"
+ }
+ },
+ "140:125": {
+ "class_type": "SamplerCustomAdvanced",
+ "inputs": {
+ "noise": [
+ "140:129",
+ 0
+ ],
+ "guider": [
+ "140:126",
+ 0
+ ],
+ "sampler": [
+ "140:123",
+ 0
+ ],
+ "sigmas": [
+ "140:124",
+ 0
+ ],
+ "latent_image": [
+ "140:131",
+ 1
+ ]
+ },
+ "_meta": {
+ "title": "SamplerCustomAdvanced"
+ }
+ },
+ "140:126": {
+ "class_type": "BasicGuider",
+ "inputs": {
+ "model": [
+ "140:127",
+ 0
+ ],
+ "conditioning": [
+ "140:131",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Basic Guider"
+ }
+ },
+ "140:127": {
+ "class_type": "UNETLoader",
+ "inputs": {
+ "unet_name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "weight_dtype": "default"
+ },
+ "_meta": {
+ "title": "Load Diffusion Model"
+ }
+ },
+ "140:128": {
+ "class_type": "CLIPLoader",
+ "inputs": {
+ "clip_name": "qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "type": "minimax",
+ "device": "default"
+ },
+ "_meta": {
+ "title": "Load CLIP"
+ }
+ },
+ "140:129": {
+ "class_type": "RandomNoise",
+ "inputs": {
+ "noise_seed": 1
+ },
+ "_meta": {
+ "title": "RandomNoise"
+ }
+ },
+ "140:130": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "140:122",
+ 0
+ ],
+ "audio": [
+ "140:121",
+ 0
+ ],
+ "fps": 24,
+ "bit_depth": 8
+ },
+ "_meta": {
+ "title": "Create Video"
+ }
+ },
+ "140:131": {
+ "class_type": "MiniMaxH3ImageToVideo",
+ "inputs": {
+ "clip": [
+ "140:128",
+ 0
+ ],
+ "vae": [
+ "140:119",
+ 0
+ ],
+ "width": [
+ "115",
+ 0
+ ],
+ "height": [
+ "115",
+ 1
+ ],
+ "length": [
+ "140:132",
+ 1
+ ],
+ "prompt": "Vaporwave title sequence look: pink and blue gradient palette, VHS tracking artifacts, Greek statue motifs, chrome palm trees, RGB chromatic aberration, lo-fi retro atmosphere, mood languid and nostalgic.\n\nTimeline:\n[0s-1s] VHS static opens the frame, the title \"COMFYUI\" appears with RGB split and a slight horizontal jitter.\n[1s-2.5s] Hard cut, a Greek plaster bust close-up, pink-purple gradient sky, a pixelated sun.\n[2.5s-4s] Clean \"STARRING\" credits appear, \"LATENT\" and \"CONTROLNET\" each shown exactly once.\n[4s-5s] Final card \"DIRECTED BY COMFYUI\" holds, one VHS tracking glitch settling into stability.\n\nHard cuts only, transitions landing with tape jumps, no push-ins, no dissolves.\n\nAudio: lo-fi vaporwave score, slow drum machine with soft bass, VHS tape-noise sample joins at 2.5s, melody fading for the last 1s.\n\nAll text must be clearly legible, do not misspell English, no Chinese characters, do not repeat names or job titles, no soft dissolves, no subtitle bars."
+ },
+ "_meta": {
+ "title": "MiniMax H3 Image to Video"
+ }
+ },
+ "140:132": {
+ "class_type": "ComfyMathExpression",
+ "inputs": {
+ "values.a": [
+ "140:133",
+ 0
+ ],
+ "expression": "max(5, round(a * 24)) + (5 - (max(5, round(a * 24)) % 17)) % 17"
+ },
+ "_meta": {
+ "title": "Math Expression"
+ }
+ },
+ "140:133": {
+ "class_type": "PrimitiveFloat",
+ "inputs": {
+ "value": 2
+ },
+ "_meta": {
+ "title": "Float (duration)"
+ }
+ },
+ "140:137": {
+ "class_type": "PrimitiveInt",
+ "inputs": {
+ "value": 20
+ },
+ "_meta": {
+ "title": "Int"
+ }
+ }
+}
diff --git a/apps/video-remove-bg/app.json b/apps/video-remove-bg/app.json
new file mode 100644
index 0000000..0dee822
--- /dev/null
+++ b/apps/video-remove-bg/app.json
@@ -0,0 +1,129 @@
+{
+ "$schema": "../../schema/app.schema.json",
+ "id": "video-remove-bg",
+ "name": "Video Background Remover",
+ "section": "video",
+ "summary": "Cuts the subject out of every frame of a clip with BiRefNet: the clip over a solid color, with its sound, and the matte for compositing.",
+ "higgsfield_feature": "Video Background Remover",
+ "pack": "upscale",
+ "status": "ready",
+ "workflow": "workflow.api.json",
+ "source": {
+ "template": "utility_birefnet_remove_background.json",
+ "comfyui": "0.42.0",
+ "notes": "BiRefNet's two nodes from Comfy's remove-background template, run on every frame with core video nodes (tools/build_apps.py: video_remove_bg)."
+ },
+ "inputs": [
+ {
+ "id": "clip",
+ "type": "video",
+ "label": "Your clip",
+ "required": true,
+ "limits": {
+ "max_mb": 100
+ },
+ "bind": [
+ {
+ "node": "1",
+ "field": "file"
+ }
+ ]
+ },
+ {
+ "id": "background",
+ "type": "select",
+ "label": "Background",
+ "default": "green",
+ "options": [
+ {
+ "value": "green",
+ "label": "Green screen",
+ "sets": [
+ {
+ "node": "8",
+ "field": "color",
+ "value": 45376
+ }
+ ]
+ },
+ {
+ "value": "white",
+ "label": "White",
+ "sets": [
+ {
+ "node": "8",
+ "field": "color",
+ "value": 16777215
+ }
+ ]
+ },
+ {
+ "value": "black",
+ "label": "Black",
+ "sets": [
+ {
+ "node": "8",
+ "field": "color",
+ "value": 0
+ }
+ ]
+ },
+ {
+ "value": "blue",
+ "label": "Blue screen",
+ "sets": [
+ {
+ "node": "8",
+ "field": "color",
+ "value": 18175
+ }
+ ]
+ }
+ ]
+ },
+ {
+ "id": "seconds",
+ "type": "slider",
+ "label": "Seconds to process",
+ "min": 1,
+ "max": 20,
+ "step": 1,
+ "default": 10,
+ "help": "From the start of the clip; every frame is cut out.",
+ "bind": [
+ {
+ "node": "2",
+ "field": "duration"
+ }
+ ]
+ }
+ ],
+ "outputs": [
+ {
+ "node": "11",
+ "type": "video",
+ "label": "Cutout"
+ },
+ {
+ "node": "14",
+ "type": "video",
+ "label": "Matte"
+ }
+ ],
+ "models": [
+ {
+ "repo": "Comfy-Org/BiRefNet",
+ "file": "background_removal/birefnet.safetensors",
+ "dir": "background_removal",
+ "gb": 0.44,
+ "phase": "all",
+ "license": "MIT"
+ }
+ ],
+ "credits": [
+ "BiRefNet (MIT)"
+ ],
+ "test": {
+ "clip": "test/clip.mp4"
+ }
+}
diff --git a/apps/video-remove-bg/workflow.api.json b/apps/video-remove-bg/workflow.api.json
new file mode 100644
index 0000000..05912c3
--- /dev/null
+++ b/apps/video-remove-bg/workflow.api.json
@@ -0,0 +1,212 @@
+{
+ "1": {
+ "class_type": "LoadVideo",
+ "inputs": {
+ "file": "clip.mp4"
+ },
+ "_meta": {
+ "title": "Clip"
+ }
+ },
+ "2": {
+ "class_type": "Video Slice",
+ "inputs": {
+ "video": [
+ "1",
+ 0
+ ],
+ "start_time": 0.0,
+ "duration": 10.0,
+ "strict_duration": false
+ },
+ "_meta": {
+ "title": "First 10 s"
+ }
+ },
+ "3": {
+ "class_type": "GetVideoComponents",
+ "inputs": {
+ "video": [
+ "2",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Frames, sound, fps"
+ }
+ },
+ "4": {
+ "class_type": "ImageScaleToTotalPixels",
+ "inputs": {
+ "image": [
+ "3",
+ 0
+ ],
+ "upscale_method": "area",
+ "megapixels": 0.9,
+ "resolution_steps": 2
+ },
+ "_meta": {
+ "title": "Frames at 0.9 MP"
+ }
+ },
+ "5": {
+ "class_type": "LoadBackgroundRemovalModel",
+ "inputs": {
+ "bg_removal_name": "birefnet.safetensors"
+ },
+ "_meta": {
+ "title": "Load Background Removal Model"
+ }
+ },
+ "6": {
+ "class_type": "RemoveBackground",
+ "inputs": {
+ "image": [
+ "4",
+ 0
+ ],
+ "bg_removal_model": [
+ "5",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Subject mask per frame"
+ }
+ },
+ "7": {
+ "class_type": "GetImageSize",
+ "inputs": {
+ "image": [
+ "4",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Size"
+ }
+ },
+ "8": {
+ "class_type": "EmptyImage",
+ "inputs": {
+ "width": [
+ "7",
+ 0
+ ],
+ "height": [
+ "7",
+ 1
+ ],
+ "batch_size": [
+ "7",
+ 2
+ ],
+ "color": 45376
+ },
+ "_meta": {
+ "title": "Background color"
+ }
+ },
+ "9": {
+ "class_type": "ImageCompositeMasked",
+ "inputs": {
+ "destination": [
+ "8",
+ 0
+ ],
+ "source": [
+ "4",
+ 0
+ ],
+ "x": 0,
+ "y": 0,
+ "resize_source": false,
+ "mask": [
+ "6",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Subject over the color"
+ }
+ },
+ "10": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "9",
+ 0
+ ],
+ "fps": [
+ "3",
+ 2
+ ],
+ "audio": [
+ "3",
+ 1
+ ]
+ },
+ "_meta": {
+ "title": "Cutout video"
+ }
+ },
+ "11": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "10",
+ 0
+ ],
+ "format": "auto",
+ "format.codec": "auto",
+ "filename_prefix": "freefield/video-remove-bg"
+ },
+ "_meta": {
+ "title": "Save cutout"
+ }
+ },
+ "12": {
+ "class_type": "MaskToImage",
+ "inputs": {
+ "mask": [
+ "6",
+ 0
+ ]
+ },
+ "_meta": {
+ "title": "Matte"
+ }
+ },
+ "13": {
+ "class_type": "CreateVideo",
+ "inputs": {
+ "images": [
+ "12",
+ 0
+ ],
+ "fps": [
+ "3",
+ 2
+ ]
+ },
+ "_meta": {
+ "title": "Matte video"
+ }
+ },
+ "14": {
+ "class_type": "SaveVideo",
+ "inputs": {
+ "video": [
+ "13",
+ 0
+ ],
+ "format": "auto",
+ "format.codec": "auto",
+ "filename_prefix": "freefield/video-remove-bg"
+ },
+ "_meta": {
+ "title": "Save matte"
+ }
+ }
+}
diff --git a/docs/import-workflow.md b/docs/import-workflow.md
new file mode 100644
index 0000000..6405550
--- /dev/null
+++ b/docs/import-workflow.md
@@ -0,0 +1,79 @@
+# Import a ComfyUI workflow as an app
+
+`tools/import_workflow.py` turns a ComfyUI workflow into a draft Freefield app: `apps/<id>/app.json` and
+`apps/<id>/workflow.api.json`. The draft passes `npm run validate`; you check what it guessed, then mark it ready.
+
+```sh
+python tools/import_workflow.py my_workflow.json --id my-app --name "My App"
+python tools/import_workflow.py ui_workflow.json --id my-app --object-info object_info.json # a UI workflow
+```
+
+The file can be in API format (*Export (API)* in ComfyUI) or a UI workflow with nodes and links. A UI workflow is
+converted with `tools/ui2api.py`; when it uses a node that `tools/object_info.json` does not list, pass
+`--object-info` with `GET /object_info` from a ComfyUI that has that node. `--section` (image, video, audio,
+utility), `--out` (another folder) and `--force` (overwrite) are optional.
+
+## What it reads from the graph
+
+| In the workflow | In app.json |
+|---|---|
+| `LoadImage`, `LoadVideo`, `LoadAudio` | a required image, video or audio input |
+| the words of a text encoder that reach a `positive` input, or the `PrimitiveString` node that feeds it | `prompt` |
+| the same for `negative`, when it has words | `negative` |
+| `seed` / `noise_seed` on samplers and noise nodes | one `seed` input that writes all of them |
+| `steps` on samplers and schedulers | a slider from half to twice the value |
+| literal `width`, `height`, `length` on an `Empty…Latent…` node | selects (768–1536) and a frames slider |
+| `ResolutionSelector` | an aspect-ratio select with the node's options |
+| `PrimitiveInt`, `PrimitiveFloat`, `PrimitiveBoolean` with a title | a slider or toggle labelled with the title |
+| `Save…` nodes (`PreviewImage` becomes `SaveImage`) | outputs, saved under `freefield/<id>` |
+| every loader value that ends in `.safetensors`, `.ckpt`, `.pt`, `.pth`, `.gguf`, `.bin` | models |
+
+A model file that one of `packs/*.json` already lists gets its repo, path, size, folder, license and gate from
+there, and its phase from the apps that use it. The pack is the existing one that holds every model; otherwise the
+app gets a new pack named after its id. A file no pack lists gets `"repo": "TODO/unknown"` and a size of 0.001 GB,
+and the report names it.
+
+## Example
+
+The official FLUX 3 Klein edit template, in UI format:
+
+```
+ui-test: 7 inputs, 1 outputs, 4 models → apps/ui-test
+ inputs:
+ - image (image) → 11.image
+ - prompt (textarea) → 48.value
+ - prompt2 (textarea) → 49.prompt
+ - steps (slider) → 7.steps
+ - width (select) → 44.width
+ - height (select) → 44.height
+ - seed (seed) → 7.seed
+ outputs:
+ - image ← 9 (SaveImage)
+ models matched in packs/:
+ - flux3-image-klein-int8-convrot.safetensors black-forest-labs/flux-3-image-klein-comfy 5.83 GB (denoise)
+ - qwen3vl_4b_int8_convrot.safetensors black-forest-labs/flux-3-image-klein-comfy 4.86 GB (encode)
+ - flux2-vae.safetensors black-forest-labs/flux-3-image-klein-comfy 0.34 GB (decode)
+ models Freefield does not know:
+ - qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors (loaded by CLIPLoader 46.clip_name; folder guessed as text_encoders)
+ check by hand:
+ - converted from the UI format with object_info.json
+ pack: ui-test (new: no existing pack holds every model)
+ validator: 0 errors, 0 warnings
+```
+
+It exits 1 when the app does not validate, 2 on a bad id or an existing folder without `--force`.
+
+## What it cannot know
+
+- Which values a user should change. Every titled primitive becomes an input, frame rates and switches included;
+ delete the ones that are internal.
+- Which seed matters. All seed fields get the same input; a two-stage graph often wants only one of them.
+- Prompts that are not literal text: words assembled by `StringConcatenate`, or a prompt-enhancer's own
+ instructions, which show up as a second prompt (`prompt2` above).
+- Where an unknown model comes from. Find the file on Hugging Face, fill `repo`, `file` and `gb`, then run
+ `python tools/catalog.py all`; until then the new pack lists a TODO repo.
+- Presets, gallery selects, templates around `{input}` values, credits and a showcase lab case: those are written
+ by hand, as in the other apps.
+
+Tests: `python test/import_workflow_test.py` imports `z-image`, `remove-bg`, `edit` and `video` and compares the
+result with their hand-written app.json.
diff --git a/docs/vram-map.json b/docs/vram-map.json
index 8e8f210..9eec17d 100644
--- a/docs/vram-map.json
+++ b/docs/vram-map.json
@@ -54,6 +54,58 @@
"measured_peak_gb": 4.96,
"measured_on": "L40S · 20261020-133620-l40s"
},
+ {
+ "id": "video-remove-bg",
+ "name": "Video Background Remover",
+ "section": "video",
+ "pack": "upscale",
+ "status": "ready",
+ "whole_gb": 0.44,
+ "phases_gb": {
+ "encode": 0.44,
+ "denoise": 0.44,
+ "denoise2": 0.0,
+ "decode": 0.44,
+ "aux": 0.0
+ },
+ "largest_phase": "encode",
+ "largest_phase_gb": 0.44,
+ "optional_gb": 0,
+ "biggest_file": {
+ "name": "birefnet.safetensors",
+ "gb": 0.44
+ },
+ "verdict_24gb": "fits whole",
+ "modal_gpu": "L4 / A10G (24 GB)",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
+ {
+ "id": "depth-map",
+ "name": "Depth Map",
+ "section": "image",
+ "pack": "upscale",
+ "status": "ready",
+ "whole_gb": 0.66,
+ "phases_gb": {
+ "encode": 0.66,
+ "denoise": 0.66,
+ "denoise2": 0.0,
+ "decode": 0.66,
+ "aux": 0.0
+ },
+ "largest_phase": "encode",
+ "largest_phase_gb": 0.66,
+ "optional_gb": 0,
+ "biggest_file": {
+ "name": "moge_2_vitl_normal_fp16.safetensors",
+ "gb": 0.66
+ },
+ "verdict_24gb": "fits whole",
+ "modal_gpu": "L4 / A10G (24 GB)",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
{
"id": "upscale",
"name": "Upscale Image",
@@ -418,6 +470,32 @@
"measured_peak_gb": 11.17,
"measured_on": "H100 · 20261022-035128-h100"
},
+ {
+ "id": "expand-image",
+ "name": "Expand Image",
+ "section": "image",
+ "pack": "image-flux3",
+ "status": "ready",
+ "whole_gb": 11.03,
+ "phases_gb": {
+ "encode": 4.86,
+ "denoise": 5.83,
+ "denoise2": 0.0,
+ "decode": 0.34,
+ "aux": 0.0
+ },
+ "largest_phase": "denoise",
+ "largest_phase_gb": 5.83,
+ "optional_gb": 0,
+ "biggest_file": {
+ "name": "flux3-image-klein-int8-convrot.safetensors",
+ "gb": 5.83
+ },
+ "verdict_24gb": "fits whole",
+ "modal_gpu": "L4 / A10G (24 GB)",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
{
"id": "face-swap",
"name": "Face Swap",
@@ -1354,6 +1432,32 @@
"measured_peak_gb": 11.18,
"measured_on": "H100 · 20261023-225832-h100"
},
+ {
+ "id": "ai-influencer",
+ "name": "AI Influencer",
+ "section": "image",
+ "pack": "image",
+ "status": "ready",
+ "whole_gb": 12.17,
+ "phases_gb": {
+ "encode": 5.63,
+ "denoise": 6.2,
+ "denoise2": 0.0,
+ "decode": 0.34,
+ "aux": 0.0
+ },
+ "largest_phase": "denoise",
+ "largest_phase_gb": 6.2,
+ "optional_gb": 0,
+ "biggest_file": {
+ "name": "z_image_turbo_int8_convrot.safetensors",
+ "gb": 6.2
+ },
+ "verdict_24gb": "fits whole",
+ "modal_gpu": "L4 / A10G (24 GB)",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
{
"id": "cinema-studio",
"name": "Cinema Studio",
@@ -1562,6 +1666,32 @@
"measured_peak_gb": 41.26,
"measured_on": "L40S · 20261020-133620-l40s"
},
+ {
+ "id": "bullet-time-scene",
+ "name": "Bullet Time Scene",
+ "section": "video",
+ "pack": "video-wan-effects",
+ "status": "ready",
+ "whole_gb": 24.65,
+ "phases_gb": {
+ "encode": 8.0,
+ "denoise": 16.4,
+ "denoise2": 0.0,
+ "decode": 0.25,
+ "aux": 0.0
+ },
+ "largest_phase": "denoise",
+ "largest_phase_gb": 16.4,
+ "optional_gb": 1.1,
+ "biggest_file": {
+ "name": "wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "gb": 16.4
+ },
+ "verdict_24gb": "fits by phases",
+ "modal_gpu": "L4 / A10G (24 GB)",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
{
"id": "motion-transfer",
"name": "Motion Transfer",
@@ -2446,6 +2576,58 @@
"measured_peak_gb": 51.25,
"measured_on": "H100 · 20261023-143551-h100"
},
+ {
+ "id": "bullet-time-splash",
+ "name": "Bullet Time Splash",
+ "section": "video",
+ "pack": "video-wan-effects",
+ "status": "ready",
+ "whole_gb": 35.68,
+ "phases_gb": {
+ "encode": 12.86,
+ "denoise": 22.23,
+ "denoise2": 0.0,
+ "decode": 0.59,
+ "aux": 0.0
+ },
+ "largest_phase": "denoise",
+ "largest_phase_gb": 22.23,
+ "optional_gb": 1.1,
+ "biggest_file": {
+ "name": "wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "gb": 16.4
+ },
+ "verdict_24gb": "tight (by phases)",
+ "modal_gpu": "test on L4 / A10G; L40S (48 GB) is comfortable",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
+ {
+ "id": "bullet-time-white",
+ "name": "Bullet Time White",
+ "section": "video",
+ "pack": "video-wan-effects",
+ "status": "ready",
+ "whole_gb": 35.68,
+ "phases_gb": {
+ "encode": 12.86,
+ "denoise": 22.23,
+ "denoise2": 0.0,
+ "decode": 0.59,
+ "aux": 0.0
+ },
+ "largest_phase": "denoise",
+ "largest_phase_gb": 22.23,
+ "optional_gb": 1.1,
+ "biggest_file": {
+ "name": "wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
+ "gb": 16.4
+ },
+ "verdict_24gb": "tight (by phases)",
+ "modal_gpu": "test on L4 / A10G; L40S (48 GB) is comfortable",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
{
"id": "vfx-water",
"name": "Water Simulation",
@@ -2576,6 +2758,84 @@
"measured_peak_gb": 43.67,
"measured_on": "H100 · 20261023-044125-h100"
},
+ {
+ "id": "extend-video",
+ "name": "Extend Video",
+ "section": "video",
+ "pack": "video-h3",
+ "status": "ready",
+ "whole_gb": 51.53,
+ "phases_gb": {
+ "encode": 27.14,
+ "denoise": 20.97,
+ "denoise2": 0.0,
+ "decode": 3.42,
+ "aux": 0.0
+ },
+ "largest_phase": "encode",
+ "largest_phase_gb": 27.14,
+ "optional_gb": 0,
+ "biggest_file": {
+ "name": "qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "gb": 27.14
+ },
+ "verdict_24gb": "needs Modal",
+ "modal_gpu": "L40S (48 GB), by phases",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
+ {
+ "id": "h3-photo",
+ "name": "H3 Photo to Video",
+ "section": "video",
+ "pack": "video-h3",
+ "status": "ready",
+ "whole_gb": 51.53,
+ "phases_gb": {
+ "encode": 27.14,
+ "denoise": 20.97,
+ "denoise2": 0.0,
+ "decode": 3.42,
+ "aux": 0.0
+ },
+ "largest_phase": "encode",
+ "largest_phase_gb": 27.14,
+ "optional_gb": 0,
+ "biggest_file": {
+ "name": "qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "gb": 27.14
+ },
+ "verdict_24gb": "needs Modal",
+ "modal_gpu": "L40S (48 GB), by phases",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
+ {
+ "id": "h3-video",
+ "name": "H3 Video",
+ "section": "video",
+ "pack": "video-h3",
+ "status": "ready",
+ "whole_gb": 51.53,
+ "phases_gb": {
+ "encode": 27.14,
+ "denoise": 20.97,
+ "denoise2": 0.0,
+ "decode": 3.42,
+ "aux": 0.0
+ },
+ "largest_phase": "encode",
+ "largest_phase_gb": 27.14,
+ "optional_gb": 0,
+ "biggest_file": {
+ "name": "qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
+ "gb": 27.14
+ },
+ "verdict_24gb": "needs Modal",
+ "modal_gpu": "L40S (48 GB), by phases",
+ "measured_peak_gb": null,
+ "measured_on": null
+ },
{
"id": "reshoot",
"name": "Re-shoot",
diff --git a/docs/vram-map.md b/docs/vram-map.md
index 1bc0eee..52b01bc 100644
--- a/docs/vram-map.md
+++ b/docs/vram-map.md
@@ -14,6 +14,8 @@ is not the least the app needs.
|---|---|---:|---|---|---|---|---:|
| `smooth` | ready | 0.1 GB | encode 0.1 GB | film_net_fp16.safetensors (0.1) | fits whole | L4 / A10G (24 GB) | 27.7 GB (L40S) |
| `remove-bg` | ready | 0.4 GB | encode 0.4 GB | birefnet.safetensors (0.4) | fits whole | L4 / A10G (24 GB) | 5.0 GB (L40S) |
+| `video-remove-bg` | ready | 0.4 GB | encode 0.4 GB | birefnet.safetensors (0.4) | fits whole | L4 / A10G (24 GB) | not measured |
+| `depth-map` | ready | 0.7 GB | encode 0.7 GB | moge_2_vitl_normal_fp16.safetensors (0.7) | fits whole | L4 / A10G (24 GB) | not measured |
| `upscale` | ready | 4.0 GB | denoise 3.5 GB | seedvr2_3b_int8_convrot.safetensors (3.5) | fits whole | L4 / A10G (24 GB) | 5.6 GB (L40S) |
| `upscale-video` | ready | 4.0 GB | denoise 3.5 GB | seedvr2_3b_int8_convrot.safetensors (3.5) | fits whole | L4 / A10G (24 GB) | 26.1 GB (L40S) |
| `billboard` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 11.9 GB (H100) |
@@ -28,6 +30,7 @@ is not the least the app needs.
| `commercial-faces` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 11.2 GB (H100) |
| `edit` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 30.9 GB (L40S) |
| `edit-text` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 11.2 GB (H100) |
+| `expand-image` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | not measured |
| `face-swap` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 12.3 GB (L40S) |
| `fashion-factory` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 11.9 GB (H100) |
| `figure-3d` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 11.2 GB (H100) |
@@ -64,6 +67,7 @@ is not the least the app needs.
| `try-on` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 11.8 GB (H100) |
| `ugc-shot` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 27.9 GB (H100) |
| `victory-card` | ready | 11.0 GB | denoise 5.8 GB | flux3-image-klein-int8-convrot.safetensors (5.8) | fits whole | L4 / A10G (24 GB) | 11.2 GB (H100) |
+| `ai-influencer` | ready | 12.2 GB | denoise 6.2 GB | z_image_turbo_int8_convrot.safetensors (6.2) | fits whole | L4 / A10G (24 GB) | not measured |
| `cinema-studio` | ready | 12.2 GB | denoise 6.2 GB | z_image_turbo_int8_convrot.safetensors (6.2) | fits whole | L4 / A10G (24 GB) | 12.4 GB (L40S) |
| `meme` | ready | 12.2 GB | denoise 6.2 GB | z_image_turbo_int8_convrot.safetensors (6.2) | fits whole | L4 / A10G (24 GB) | 12.7 GB (H100) |
| `z-image` | ready | 12.2 GB | denoise 6.2 GB | z_image_turbo_int8_convrot.safetensors (6.2) | fits whole | L4 / A10G (24 GB) | 12.2 GB (L40S) |
@@ -72,6 +76,7 @@ is not the least the app needs.
| `style-reference` | ready | 19.4 GB | denoise 13.9 GB | krea2_turbo_int8_convrot.safetensors (13.5) | fits whole | L4 / A10G (24 GB) | 19.8 GB (L40S) |
| `styles` | ready | 19.4 GB (+3.8 optional) | denoise 14.0 GB | krea2_turbo_int8_convrot.safetensors (13.5) | fits whole | L4 / A10G (24 GB) | 19.5 GB (L40S) |
| `camera-moves` | ready | 37.6 GB (+2.5 optional) | denoise 15.3 GB | wan2.2_fun_camera_high_noise_14B_fp8_scaled.safetensors (15.3) | fits by phases | L4 / A10G (24 GB) | 41.3 GB (L40S) |
+| `bullet-time-scene` | ready | 24.6 GB (+1.1 optional) | denoise 16.4 GB | wan2.1_i2v_480p_14B_fp8_scaled.safetensors (16.4) | fits by phases | L4 / A10G (24 GB) | not measured |
| `motion-transfer` | ready | 24.9 GB | denoise 16.6 GB | wan_animate_2_distill_int8_convrot.safetensors (16.6) | fits by phases | L4 / A10G (24 GB) | 37.3 GB (L40S) |
| `effects` | ready | 25.0 GB (+9.0 optional) | denoise 16.8 GB | wan2.1_i2v_480p_14B_fp8_scaled.safetensors (16.4) | fits by phases | L4 / A10G (24 GB) | 41.9 GB (H100) |
| `recast` | ready | 28.6 GB | denoise 18.6 GB | wan2.1_14B_SCAIL_2_int8_convrot.safetensors (16.6) | fits by phases | L4 / A10G (24 GB) | 30.6 GB (L40S) |
@@ -106,11 +111,16 @@ is not the least the app needs.
| `volcano-ad` | ready | 50.1 GB (+5.2 optional) | denoise 21.9 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | 50.3 GB (H100) |
| `camera-copy` | ready | 39.4 GB (+5.2 optional) | denoise 22.1 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | 40.6 GB (L40S) |
| `mixed-media` | ready | 51.4 GB | denoise 22.1 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | 51.2 GB (H100) |
+| `bullet-time-splash` | ready | 35.7 GB (+1.1 optional) | denoise 22.2 GB | wan2.1_i2v_480p_14B_fp8_scaled.safetensors (16.4) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | not measured |
+| `bullet-time-white` | ready | 35.7 GB (+1.1 optional) | denoise 22.2 GB | wan2.1_i2v_480p_14B_fp8_scaled.safetensors (16.4) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | not measured |
| `vfx-water` | ready | 39.6 GB (+5.2 optional) | denoise 22.4 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | 44.4 GB (L40S) |
| `genjutsu` | ready | 40.0 GB | denoise 22.8 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | 40.9 GB (H100 NVL) |
| `reframe` | ready | 40.0 GB | denoise 22.8 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | 44.1 GB (H100) |
| `sdr-hdr` | ready | 24.6 GB | denoise 22.8 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | tight (by phases) | test on L4 / A10G; L40S (48 GB) is comfortable | 37.3 GB (H100 NVL) |
| `dub` | ready | 41.2 GB (+5.2 optional) | denoise 24.0 GB | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors (21.5) | needs Modal | L40S (48 GB), whole | 43.7 GB (H100) |
+| `extend-video` | ready | 51.5 GB | encode 27.1 GB | qwen3vl_32b_minimax_h3_int8_convrot.safetensors (27.1) | needs Modal | L40S (48 GB), by phases | not measured |
+| `h3-photo` | ready | 51.5 GB | encode 27.1 GB | qwen3vl_32b_minimax_h3_int8_convrot.safetensors (27.1) | needs Modal | L40S (48 GB), by phases | not measured |
+| `h3-video` | ready | 51.5 GB | encode 27.1 GB | qwen3vl_32b_minimax_h3_int8_convrot.safetensors (27.1) | needs Modal | L40S (48 GB), by phases | not measured |
| `reshoot` | ready | 54.8 GB | encode 27.1 GB | qwen3vl_32b_minimax_h3_int8_convrot.safetensors (27.1) | needs Modal | L40S (48 GB), by phases | 44.1 GB (L40S) |
| `smoke` | ready | 0.0 GB | encode 0.0 GB | — | no models | any | not measured |
| `smoke-invert` | ready | 0.0 GB | encode 0.0 GB | — | no models | any | not measured |
diff --git a/lab/cases.json b/lab/cases.json
index 369492b..daeb1e0 100644
--- a/lab/cases.json
+++ b/lab/cases.json
@@ -437,6 +437,12 @@
{"id": "refx-woman-b-1", "app": "z-image", "values": {"prompt": "Waist-up portrait photo of a Black woman in her late thirties with short natural coily hair, wearing a navy blazer, calm expression, looking at the camera, natural skin texture with visible pores, no retouching, soft studio light, plain light blue background, realistic photo", "width": 768, "height": 1280}, "note": "Candidate for ref-fx-woman-b (laughing, money rain, four morph transformations, outfit shot). Chosen for ref-fx-woman-b on 24 Oct (run 20261024-164414-h100), with a white t-shirt added under the blazer: this one showed bare skin under it."},
{"id": "refx-woman-b-2", "app": "z-image", "values": {"prompt": "Waist-up portrait photo of a Lebanese woman in her late thirties with shoulder-length dark curly hair, wearing a navy blazer, calm expression, looking at the camera, natural skin texture with visible pores, no retouching, soft studio light, plain light blue background, realistic photo", "width": 768, "height": 1280}, "note": "Candidate for ref-fx-woman-b, second look. Not chosen (run 20261024-164414-h100); also bare skin under the blazer."},
+ {"id": "depth-room", "showcase": true, "app": "depth-map", "values": {"image": "@case:ref-room", "map": "depth_colored", "detail": 6}, "note": "MoGe-2 depth of the library interior: near shelves warm, the far wall cool, no holes."},
+ {"id": "vbg-walk", "showcase": true, "app": "video-remove-bg", "values": {"clip": "@case:ref-walk", "background": "green", "seconds": 5}, "note": "The walking woman on green, hair edges clean, sound kept; the matte white on black."},
+ {"id": "h3-video-tacos", "showcase": true, "app": "h3-video", "values": {"prompt": "A street cook in Mexico City flips tacos on a hot griddle, steam and sizzling sound, he looks up and says: \"Two more coming right up!\"", "aspect": "16:9 (Widescreen)", "duration": 4}, "note": "Speech, sizzle and picture from one model: does the line come out clearly and in sync?"},
+ {"id": "h3-photo-man", "showcase": true, "app": "h3-photo", "values": {"image": "@case:ref-fx-man-b", "prompt": "The man laughs, looks at the camera and says: \"Hey, good to see you.\"", "duration": 4}, "note": "Same face as the photo, the line spoken with lip sync?"},
+ {"id": "extend-stroll", "showcase": true, "app": "extend-video", "values": {"clip": "@case:ref-stroll", "prompt": "She keeps walking along the path, slows down and looks up at the trees; birds sing and leaves rustle.", "duration": 3}, "note": "Does the new part start where the clip ends (same woman, dress, park) with no jump?"},
+
{"id": "ref-tpl-sketch", "app": "z-image", "values": {"prompt": "A rough pencil sketch on white paper of a cat sitting on a windowsill next to a potted plant, loose lines, no color", "width": 1024, "height": 1024},
"note": "Reference sketch for Sketch to Real."},
{"id": "ref-tpl-sign", "app": "z-image", "values": {"prompt": "Photo of a small bakery storefront with a wooden sign above the door that reads OPEN in bold white letters, daylight, photorealistic", "width": 1280, "height": 768},
@@ -568,6 +574,16 @@
"note": "Rap God. Performance energy; a beat plays."},
{"id": "tpl-idol", "showcase": true, "app": "idol", "values": {"image": "@case:ref-fx-woman-a", "aspect": "9:16 (Portrait Widescreen)"},
"note": "Idol. Stage, confetti, pose."},
+ {"id": "tpl-expand-image", "showcase": true, "app": "expand-image", "values": {"image": "@case:ref-fx-lake", "direction": "out"},
+ "note": "Expand Image. Does the border continue the lake, rock and sky, with no seam or gray left?"},
+ {"id": "tpl-ai-influencer", "showcase": true, "app": "ai-influencer", "values": {"niche": "fitness influencer stretching on a rooftop at sunrise in workout clothes"},
+ "note": "Influencer IA. A believable person, not a glossy render?"},
+ {"id": "tpl-bullet-time-scene", "showcase": true, "app": "bullet-time-scene", "values": {"image": "@case:ref-fx-jump", "subject": "a parkour athlete in a black t-shirt leaping between two rooftops"},
+ "note": "Bullet Time Scene. Does the camera orbit while the athlete and the city stay frozen?"},
+ {"id": "tpl-bullet-time-white", "showcase": true, "app": "bullet-time-white", "values": {"image": "@case:ref-character", "subject": "a man in a red jacket, jeans and white sneakers"},
+ "note": "Bullet Time White. White backdrop, same person and outfit, smooth orbit?"},
+ {"id": "tpl-bullet-time-splash", "showcase": true, "app": "bullet-time-splash", "values": {"image": "@case:ref-fx-runner", "subject": "a sprinter in a red and white running outfit"},
+ "note": "Bullet Time Splash. Do the drops stay frozen in the air while the camera moves?"},
{"id": "tpl-angles", "showcase": true, "app": "angles", "values": {"image": "@case:ref-fx-cafe"},
"note": "Angles 2.0. The cafe seen from 45° to the left; same people and table."},
{"id": "tpl-shots", "showcase": true, "app": "shots", "values": {"image": "@case:ref-fx-lake", "shot": "将镜头转为特写镜头 Turn the camera to a close-up."},
diff --git a/packs/image-flux3.json b/packs/image-flux3.json
index ce92c75..ba4b4b7 100644
--- a/packs/image-flux3.json
+++ b/packs/image-flux3.json
@@ -13,6 +13,7 @@
"commercial-faces",
"edit",
"edit-text",
+ "expand-image",
"face-swap",
"fashion-factory",
"figure-3d",
@@ -87,6 +88,7 @@
"commercial-faces",
"edit",
"edit-text",
+ "expand-image",
"face-swap",
"fashion-factory",
"figure-3d",
@@ -146,6 +148,7 @@
"commercial-faces",
"edit",
"edit-text",
+ "expand-image",
"face-swap",
"fashion-factory",
"figure-3d",
@@ -205,6 +208,7 @@
"commercial-faces",
"edit",
"edit-text",
+ "expand-image",
"face-swap",
"fashion-factory",
"figure-3d",
diff --git a/packs/image.json b/packs/image.json
index 96b8aab..f29b005 100644
--- a/packs/image.json
+++ b/packs/image.json
@@ -1,6 +1,7 @@
{
"pack": "image",
"apps": [
+ "ai-influencer",
"cinema-studio",
"meme",
"z-image"
@@ -18,6 +19,7 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "ai-influencer",
"cinema-studio",
"meme",
"z-image"
@@ -32,6 +34,7 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "ai-influencer",
"cinema-studio",
"meme",
"z-image"
@@ -46,6 +49,7 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "ai-influencer",
"cinema-studio",
"meme",
"z-image"
diff --git a/packs/index.json b/packs/index.json
index 1338488..d03e09c 100644
--- a/packs/index.json
+++ b/packs/index.json
@@ -17,6 +17,7 @@
},
"image": {
"apps": [
+ "ai-influencer",
"cinema-studio",
"meme",
"z-image"
@@ -38,6 +39,7 @@
"commercial-faces",
"edit",
"edit-text",
+ "expand-image",
"face-swap",
"fashion-factory",
"figure-3d",
@@ -99,19 +101,24 @@
},
"upscale": {
"apps": [
+ "depth-map",
"remove-bg",
"smooth",
"upscale",
- "upscale-video"
+ "upscale-video",
+ "video-remove-bg"
],
- "download_gb": 4.47,
+ "download_gb": 5.13,
"needs_hf_token": false
},
"video-h3": {
"apps": [
+ "extend-video",
+ "h3-photo",
+ "h3-video",
"reshoot"
],
- "download_gb": 54.75,
+ "download_gb": 75.72,
"needs_hf_token": false
},
"video-ltx25": {
@@ -174,9 +181,12 @@
},
"video-wan-effects": {
"apps": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
],
- "download_gb": 34.03,
- "needs_hf_token": false
+ "download_gb": 45.06,
+ "needs_hf_token": true
}
}
diff --git a/packs/upscale.json b/packs/upscale.json
index bc8864e..9c25f55 100644
--- a/packs/upscale.json
+++ b/packs/upscale.json
@@ -1,12 +1,14 @@
{
"pack": "upscale",
"apps": [
+ "depth-map",
"remove-bg",
"smooth",
"upscale",
- "upscale-video"
+ "upscale-video",
+ "video-remove-bg"
],
- "download_gb": 4.47,
+ "download_gb": 5.13,
"needs_hf_token": false,
"custom_nodes": [],
"models": [
@@ -19,7 +21,8 @@
"gated": false,
"license": "MIT",
"used_by": [
- "remove-bg"
+ "remove-bg",
+ "video-remove-bg"
]
},
{
@@ -47,6 +50,18 @@
"smooth"
]
},
+ {
+ "repo": "Comfy-Org/MoGe",
+ "file": "geometry_estimation/moge_2_vitl_normal_fp16.safetensors",
+ "dir": "geometry_estimation",
+ "gb": 0.66,
+ "name": "moge_2_vitl_normal_fp16.safetensors",
+ "gated": false,
+ "license": "MIT",
+ "used_by": [
+ "depth-map"
+ ]
+ },
{
"repo": "Comfy-Org/SeedVR2",
"file": "vae/seedvr2_ema_vae_fp16.safetensors",
diff --git a/packs/video-h3.json b/packs/video-h3.json
index 804fb66..171c40d 100644
--- a/packs/video-h3.json
+++ b/packs/video-h3.json
@@ -1,9 +1,12 @@
{
"pack": "video-h3",
"apps": [
+ "extend-video",
+ "h3-photo",
+ "h3-video",
"reshoot"
],
- "download_gb": 54.75,
+ "download_gb": 75.72,
"needs_hf_token": false,
"custom_nodes": [
{
@@ -12,6 +15,20 @@
}
],
"models": [
+ {
+ "repo": "Comfy-Org/MiniMax-H3",
+ "file": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 20.97,
+ "name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+ "gated": false,
+ "license": "MiniMax H3 Community License",
+ "used_by": [
+ "extend-video",
+ "h3-photo",
+ "h3-video"
+ ]
+ },
{
"repo": "Comfy-Org/MiniMax-H3",
"file": "diffusion_models/minimax_h3_ref2va_pruned_int8_convrot.safetensors",
@@ -69,6 +86,9 @@
"gated": false,
"license": "MiniMax H3 Community License",
"used_by": [
+ "extend-video",
+ "h3-photo",
+ "h3-video",
"reshoot"
]
},
@@ -81,6 +101,9 @@
"gated": false,
"license": "MiniMax H3 Community License",
"used_by": [
+ "extend-video",
+ "h3-photo",
+ "h3-video",
"reshoot"
]
},
@@ -93,6 +116,9 @@
"gated": false,
"license": "MiniMax H3 Community License",
"used_by": [
+ "extend-video",
+ "h3-photo",
+ "h3-video",
"reshoot"
]
}
diff --git a/packs/video-wan-effects.json b/packs/video-wan-effects.json
index 559c0f7..2da0164 100644
--- a/packs/video-wan-effects.json
+++ b/packs/video-wan-effects.json
@@ -1,10 +1,13 @@
{
"pack": "video-wan-effects",
"apps": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
],
- "download_gb": 34.03,
- "needs_hf_token": false,
+ "download_gb": 45.06,
+ "needs_hf_token": true,
"custom_nodes": [],
"models": [
{
@@ -16,9 +19,25 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
]
},
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "diffusion_models/flux3-image-klein-int8-convrot.safetensors",
+ "dir": "diffusion_models",
+ "gb": 5.83,
+ "name": "flux3-image-klein-int8-convrot.safetensors",
+ "gated": true,
+ "license": "FLUX Kommunity License",
+ "used_by": [
+ "bullet-time-splash",
+ "bullet-time-white"
+ ]
+ },
{
"repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
"file": "split_files/diffusion_models/wan2.1_i2v_480p_14B_fp8_scaled.safetensors",
@@ -28,6 +47,9 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
]
},
@@ -76,6 +98,9 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
]
},
@@ -256,6 +281,9 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
]
},
@@ -331,6 +359,19 @@
"effects"
]
},
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "text_encoders/qwen3vl_4b_int8_convrot.safetensors",
+ "dir": "text_encoders",
+ "gb": 4.86,
+ "name": "qwen3vl_4b_int8_convrot.safetensors",
+ "gated": true,
+ "license": "FLUX Kommunity License",
+ "used_by": [
+ "bullet-time-splash",
+ "bullet-time-white"
+ ]
+ },
{
"repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
"file": "split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors",
@@ -340,9 +381,25 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
]
},
+ {
+ "repo": "black-forest-labs/flux-3-image-klein-comfy",
+ "file": "vae/flux2-vae.safetensors",
+ "dir": "vae",
+ "gb": 0.34,
+ "name": "flux2-vae.safetensors",
+ "gated": true,
+ "license": "FLUX Kommunity License",
+ "used_by": [
+ "bullet-time-splash",
+ "bullet-time-white"
+ ]
+ },
{
"repo": "Comfy-Org/Wan_2.1_ComfyUI_repackaged",
"file": "split_files/vae/wan_2.1_vae.safetensors",
@@ -352,6 +409,9 @@
"gated": false,
"license": "Apache-2.0",
"used_by": [
+ "bullet-time-scene",
+ "bullet-time-splash",
+ "bullet-time-white",
"effects"
]
}
diff --git a/test/import_workflow_test.py b/test/import_workflow_test.py
new file mode 100644
index 0000000..c0bc34d
--- /dev/null
+++ b/test/import_workflow_test.py
@@ -0,0 +1,106 @@
+"""tools/import_workflow.py against workflows whose app.json was written by hand.
+
+ python test/import_workflow_test.py
+"""
+import json
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parent.parent
+sys.path.insert(0, str(ROOT / "tools"))
+import import_workflow as iw # tools/import_workflow.py
+
+failures = []
+
+
+def check(cond, what):
+ print(("ok " if cond else "FAIL ") + what)
+ if not cond:
+ failures.append(what)
+
+
+def binds(app):
+ return {(b["node"], b["field"]) for i in app["inputs"] for b in i.get("bind", [])}
+
+
+def kinds(app):
+ return {(b["node"], b["field"]): i["type"] for i in app["inputs"] for b in i.get("bind", [])}
+
+
+def files(app, required_only=False):
+ return {m["file"].rsplit("/", 1)[-1] for m in app["models"] if not (required_only and m.get("optional"))}
+
+
+def imported(app_id):
+ hand = json.loads((ROOT / "apps" / app_id / "app.json").read_text(encoding="utf-8"))
+ api = json.loads((ROOT / "apps" / app_id / "workflow.api.json").read_text(encoding="utf-8"))
+ app, out_api, report = iw.build(api, f"imp-{app_id}", None, None, "workflow.api.json")
+ return hand, app, out_api, report
+
+
+# the same inputs, outputs and models as the hand-written app
+for app_id in ("z-image", "remove-bg", "edit"):
+ hand, app, api, report = imported(app_id)
+ check(binds(app) == binds(hand), f"{app_id}: inputs write the same node fields ({sorted(binds(app))})")
+ check(all(kinds(app)[k] == t for k, t in kinds(hand).items()), f"{app_id}: each input has the hand-written type")
+ check({o["node"] for o in app["outputs"]} == {o["node"] for o in hand["outputs"]}, f"{app_id}: same output nodes")
+ check(files(app) == files(hand), f"{app_id}: same model files")
+ check(app["pack"] == hand["pack"], f"{app_id}: pack {app['pack']} (hand-written: {hand['pack']})")
+ check(all(m["repo"] == h["repo"] and m["dir"] == h["dir"] and m["phase"] == h["phase"]
+ for m in app["models"] for h in hand["models"] if m["file"] == h["file"]), f"{app_id}: repo, folder and phase match")
+ check(app["status"] == "draft" and not report["unknown"], f"{app_id}: a draft with no unknown model")
+ check(all(n["inputs"]["filename_prefix"] == f"freefield/imp-{app_id}" for n in api.values() if "filename_prefix" in n["inputs"]),
+ f"{app_id}: outputs saved under freefield/imp-{app_id}")
+
+# a bigger graph (LTX-2.5 image to video): everything the hand-written app exposes, plus what it chose to hide
+hand, app, api, report = imported("video")
+check(binds(hand) <= binds(app), "video: every hand-written binding is found")
+check(files(app) == files(hand), "video: same model files (the optional prompt enhancer included)")
+check({o["node"] for o in app["outputs"]} == {"75"}, "video: the SaveVideo output")
+check(app["section"] == "video" and app["pack"] == "video-ltx25", "video: section and pack from the outputs and models")
+check(any("seed writes 2 nodes" in c for c in report["check"]), "video: the report says the seed has two targets")
+check(next(i for i in app["inputs"] if i["id"] == "prompt")["bind"] == [{"node": "398:376", "field": "value"}],
+ "video: the prompt is the PrimitiveStringMultiline upstream of the switch")
+
+# a file Freefield does not know is marked, never given a made-up repo
+api = json.loads((ROOT / "apps" / "z-image" / "workflow.api.json").read_text(encoding="utf-8"))
+api["57:28"]["inputs"]["unet_name"] = "someone_elses_model.safetensors"
+app, _, report = iw.build(api, "imp-unknown", None, None, "x.json")
+m = next(m for m in app["models"] if m["file"] == "someone_elses_model.safetensors")
+check(m["repo"] == "TODO/unknown" and m["dir"] == "diffusion_models", "unknown file: TODO repo, folder from the loader")
+check(report["unknown"] and app["pack"] == "imp-unknown", "unknown file: reported, and the app gets its own pack")
+
+# a UI workflow (nodes + links) goes through tools/ui2api.py
+ui = {"nodes": [
+ {"id": 1, "type": "LoadImage", "mode": 0, "inputs": [], "widgets_values": ["photo.png", "image"],
+ "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [1]}, {"name": "MASK", "type": "MASK", "links": []}]},
+ {"id": 2, "type": "PreviewImage", "mode": 0, "inputs": [{"name": "images", "type": "IMAGE", "link": 1}], "widgets_values": []},
+], "links": [[1, 1, 0, 2, 0, "IMAGE"]]}
+with tempfile.TemporaryDirectory() as tmp:
+ p = Path(tmp) / "ui.json"
+ p.write_text(json.dumps(ui), encoding="utf-8")
+ api, notes = iw.load_workflow(p, None)
+ app, api, report = iw.build(api, "imp-ui", None, None, "ui.json")
+ check(api["2"]["class_type"] == "SaveImage" and app["outputs"] == [{"node": "2", "type": "image", "label": "Image"}],
+ "UI workflow: converted, PreviewImage became the output")
+ check(binds(app) == {("1", "image")}, "UI workflow: the LoadImage is the image input")
+
+ # the command end to end: files written, validator run on the new app alone
+ out = Path(tmp) / "app"
+ r = subprocess.run([sys.executable, str(ROOT / "tools" / "import_workflow.py"), str(ROOT / "apps" / "edit" / "workflow.api.json"),
+ "--id", "imp-edit", "--name", "Imported edit", "--out", str(out)], capture_output=True, text=True)
+ check(r.returncode == 0, f"command exits 0 ({r.stdout.splitlines()[-1] if r.stdout else r.stderr.strip()})")
+ written = json.loads((out / "app.json").read_text(encoding="utf-8"))
+ check(written["name"] == "Imported edit" and (out / "workflow.api.json").exists(), "command writes app.json and workflow.api.json")
+ checked = iw.validate(written, json.loads((out / "workflow.api.json").read_text(encoding="utf-8")))
+ check(checked is None or not checked[0], f"the written app validates ({checked})")
+ r = subprocess.run([sys.executable, str(ROOT / "tools" / "import_workflow.py"), str(ROOT / "apps" / "edit" / "workflow.api.json"),
+ "--id", "imp-edit", "--out", str(out)], capture_output=True, text=True)
+ check(r.returncode == 2, "an existing folder is not overwritten without --force")
+ r = subprocess.run([sys.executable, str(ROOT / "tools" / "import_workflow.py"), str(p), "--id", "Bad_Id"], capture_output=True, text=True)
+ check(r.returncode == 2, "a bad id is refused")
+
+print(f"\n{len(failures)} failed" if failures else "\nall passed")
+sys.exit(1 if failures else 0)
diff --git a/tools/build_apps.py b/tools/build_apps.py
index eb776e1..d927e6c 100644
--- a/tools/build_apps.py
+++ b/tools/build_apps.py
@@ -101,6 +101,85 @@ def prefix_outputs(api: Api, app_id: str) -> None:
node["inputs"]["filename_prefix"] = f"freefield/{app_id}"
+# ---------------------------------------------------------------- MiniMax H3 specifics
+def h3_local(api: Api) -> None:
+ """Comfy's MiniMax H3 templates on Freefield's weights, without the turbo LoRA.
+
+ The templates load an NVFP4 text encoder (it needs a Blackwell GPU); the int8 one reshoot already uses goes
+ in its place. Their turbo switch (a PrimitiveBoolean, off by default) picks lightx2v's 8-step LoRA or the
+ plain 20 steps; the plain branch is wired straight in and the LoRA goes, since its model card states no
+ license."""
+ swap_files(api, {"qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors": "qwen3vl_32b_minimax_h3_int8_convrot.safetensors"})
+ turbo = find(api, "PrimitiveBoolean")
+ switches = [i for i, n in api.items() if n["class_type"] == "ComfySwitchNode" and n["inputs"].get("switch") == [turbo, 0]]
+ assert len(switches) == 2, switches
+ for sw in switches:
+ rewire(api, [sw, 0], api[sw]["inputs"]["on_false"])
+ fast_steps = [api[sw]["inputs"]["on_true"][0] for sw in switches if api[api[sw]["inputs"]["on_true"][0]]["class_type"] == "PrimitiveInt"]
+ drop(api, *switches, turbo, find(api, "LoraLoaderModelOnly"), *fast_steps)
+
+
+def h3_from_photo(api: Api, megapixels: float) -> None:
+ """Size the H3 clip from the photo (the template's unused scale + size nodes) instead of a fixed aspect."""
+ scale, size = find(api, "ImageScaleToTotalPixels"), find(api, "GetImageSize")
+ photo, h3 = find(api, "LoadImage"), find(api, "MiniMaxH3ImageToVideo")
+ put(api, scale, "image", [photo, 0])
+ put(api, scale, "megapixels", megapixels)
+ put(api, scale, "upscale_method", "lanczos")
+ put(api, h3, "first_frame", [scale, 0])
+ put(api, h3, "width", [size, 0])
+ put(api, h3, "height", [size, 1])
+ drop(api, find(api, "ResolutionSelector"))
+
+
+def h3_extend(api: Api) -> None:
+ """H3 image-to-video from the clip's last frame, joined after the clip: frames with ImageBatch (it scales the
+ new frames to the clip's size), sound with AudioConcat, at the clip's frame rate."""
+ photo = find(api, "LoadImage")
+ api["ff_clip"] = {"class_type": "LoadVideo", "inputs": {"file": "clip.mp4"}, "_meta": {"title": "Clip to extend"}}
+ api["ff_parts"] = {"class_type": "GetVideoComponents", "inputs": {"video": ["ff_clip", 0]}, "_meta": {"title": "Frames, sound, fps"}}
+ api["ff_last"] = {"class_type": "ImageFromBatch", "inputs": {"image": ["ff_parts", 0], "batch_index": -1, "length": 1},
+ "_meta": {"title": "Last frame"}}
+ rewire(api, [photo, 0], ["ff_last", 0])
+ drop(api, photo)
+ decode = find(api, "VAEDecode")
+ audio = find(api, "VAEDecodeAudio")
+ out = find(api, "CreateVideo")
+ api["ff_frames"] = {"class_type": "ImageBatch", "inputs": {"image1": ["ff_parts", 0], "image2": [decode, 0]}, "_meta": {"title": "Clip + continuation"}}
+ api["ff_sound"] = {"class_type": "AudioConcat", "inputs": {"audio1": ["ff_parts", 1], "audio2": [audio, 0], "direction": "after"},
+ "_meta": {"title": "Sound + new sound"}}
+ put(api, out, "images", ["ff_frames", 0])
+ put(api, out, "audio", ["ff_sound", 0])
+ put(api, out, "fps", ["ff_parts", 2])
+
+
+def video_remove_bg(birefnet: Api) -> Api:
+ """BiRefNet (the remove-bg template's two nodes) on every frame of a clip, from core video nodes: the cutout
+ over a solid color with the clip's sound, and the matte (white subject on black) for compositing."""
+ loader = birefnet[find(birefnet, "LoadBackgroundRemovalModel")]
+ save = {"format": "auto", "format.codec": "auto", "filename_prefix": "freefield/video-remove-bg"}
+ return {
+ "1": {"class_type": "LoadVideo", "inputs": {"file": "clip.mp4"}, "_meta": {"title": "Clip"}},
+ "2": {"class_type": "Video Slice", "inputs": {"video": ["1", 0], "start_time": 0.0, "duration": 10.0, "strict_duration": False},
+ "_meta": {"title": "First 10 s"}},
+ "3": {"class_type": "GetVideoComponents", "inputs": {"video": ["2", 0]}, "_meta": {"title": "Frames, sound, fps"}},
+ "4": {"class_type": "ImageScaleToTotalPixels", "inputs": {"image": ["3", 0], "upscale_method": "area", "megapixels": 0.9, "resolution_steps": 2},
+ "_meta": {"title": "Frames at 0.9 MP"}},
+ "5": json.loads(json.dumps(loader)),
+ "6": {"class_type": "RemoveBackground", "inputs": {"image": ["4", 0], "bg_removal_model": ["5", 0]}, "_meta": {"title": "Subject mask per frame"}},
+ "7": {"class_type": "GetImageSize", "inputs": {"image": ["4", 0]}, "_meta": {"title": "Size"}},
+ "8": {"class_type": "EmptyImage", "inputs": {"width": ["7", 0], "height": ["7", 1], "batch_size": ["7", 2], "color": 45376},
+ "_meta": {"title": "Background color"}},
+ "9": {"class_type": "ImageCompositeMasked", "inputs": {"destination": ["8", 0], "source": ["4", 0], "x": 0, "y": 0, "resize_source": False,
+ "mask": ["6", 0]}, "_meta": {"title": "Subject over the color"}},
+ "10": {"class_type": "CreateVideo", "inputs": {"images": ["9", 0], "fps": ["3", 2], "audio": ["3", 1]}, "_meta": {"title": "Cutout video"}},
+ "11": {"class_type": "SaveVideo", "inputs": {"video": ["10", 0], **save}, "_meta": {"title": "Save cutout"}},
+ "12": {"class_type": "MaskToImage", "inputs": {"mask": ["6", 0]}, "_meta": {"title": "Matte"}},
+ "13": {"class_type": "CreateVideo", "inputs": {"images": ["12", 0], "fps": ["3", 2]}, "_meta": {"title": "Matte video"}},
+ "14": {"class_type": "SaveVideo", "inputs": {"video": ["13", 0], **save}, "_meta": {"title": "Save matte"}},
+ }
+
+
# ---------------------------------------------------------------- LTX specifics
LTX_INT8 = {
"ltx-2.5-22b-distilled-transformer-bf16.safetensors": "ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors",
@@ -578,6 +657,28 @@ def recipes(templates: Path, ltx: Path, info: dict, reshoot: Path | None = None)
out["music"] = tpl("audio_minimax_music_3.json")
out["sfx"] = tpl("audio_stable_audio_3_medium.json")
out["motion-transfer"] = tpl("video_wan_animate2_distilled.json")
+ dm = tpl("utility_moge_depth_estimation.json")
+ # MoGe-2 (the geometry model reshoot loads too): one render saved, the map picked in the app; the depth and
+ # mask previews go
+ colored = find(dm, "MoGeRender", output="depth_colored")
+ to_save_image(dm, find(dm, "PreviewImage", images=[colored, 0]), "freefield/depth-map")
+ drop(dm, find(dm, "PreviewImage"), find(dm, "MaskPreview"))
+ drop(dm, find(dm, "ImageToMask"))
+ drop(dm, find(dm, "MoGeRender", output="depth"), find(dm, "MoGeRender", output="mask"))
+ out["depth-map"] = dm
+ out["video-remove-bg"] = video_remove_bg(tpl("utility_birefnet_remove_background.json"))
+ h3t = tpl("video_minimax_h3_t2v.json")
+ h3_local(h3t)
+ out["h3-video"] = h3t
+ h3i = tpl("video_minimax_h3_i2v.json")
+ h3_local(h3i)
+ h3_from_photo(h3i, 0.6)
+ out["h3-photo"] = h3i
+ ext = tpl("video_minimax_h3_i2v.json")
+ h3_local(ext)
+ h3_from_photo(ext, 0.6)
+ h3_extend(ext)
+ out["extend-video"] = ext
if reshoot:
# Comfy's own reference app (Comfy-Org/comfy-reshoot, MIT), already in API format
out["reshoot"] = json.loads((reshoot / "generate.api.json").read_text(encoding="utf-8"))
diff --git a/tools/import_workflow.py b/tools/import_workflow.py
new file mode 100644
index 0000000..821ede2
--- /dev/null
+++ b/tools/import_workflow.py
@@ -0,0 +1,448 @@
+#!/usr/bin/env python3
+"""Turn a ComfyUI workflow into a draft Freefield app: apps/<id>/app.json + apps/<id>/workflow.api.json.
+
+ python tools/import_workflow.py my_workflow.json --id my-app [--name "My App"] [--section image]
+ python tools/import_workflow.py ui_workflow.json --id my-app --object-info object_info.json # a UI workflow
+ python tools/import_workflow.py wf.json --id my-app --out /tmp/my-app --force # elsewhere, overwrite
+
+The workflow may be in API format ("Export (API)") or a UI workflow with nodes and links (converted with
+tools/ui2api.py; pass --object-info with GET /object_info from a ComfyUI that has its nodes when the bundled
+tools/object_info.json lacks one). What it writes:
+
+- inputs: each LoadImage / LoadVideo / LoadAudio, the positive prompt (the literal text of a text encoder, or the
+ PrimitiveString node that feeds it), the negative prompt when it has words, seeds, steps, literal latent
+ width/height, a ResolutionSelector aspect ratio, and Primitive nodes someone gave a title;
+- outputs: every Save node (a PreviewImage becomes a SaveImage), filename_prefix "freefield/<id>";
+- models: every loader file, with repo, path, size, folder and license taken from packs/*.json when Freefield
+ already knows the file, and its phase from the apps that use it. A file Freefield does not know gets
+ repo "TODO/unknown" and a note: find it on Hugging Face before the app is used;
+- status "draft", and the pack that already holds all its models, when one does.
+
+Then it runs the repo's validator on the new app alone and prints a report. It exits 1 when the app does not
+validate (or would not), 2 on bad arguments. Nothing in the workflow is executed.
+"""
+from __future__ import annotations
+
+import argparse
+import json
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+from collections import Counter, defaultdict
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parent.parent
+sys.path.insert(0, str(Path(__file__).resolve().parent))
+from ui2api import ConversionError, convert # noqa: E402 tools/ui2api.py
+
+MODEL_EXT = (".safetensors", ".ckpt", ".pt", ".pth", ".gguf", ".bin", ".sft")
+FILE_LOADERS = {"LoadImage": ("image", "image"), "LoadImageMask": ("image", "image"),
+ "LoadVideo": ("video", "file"), "LoadAudio": ("audio", "audio")}
+SAVE_TYPES = {"SaveImage": "image", "SaveAnimatedWEBP": "image", "SaveAnimatedPNG": "image",
+ "SaveVideo": "video", "SaveWEBM": "video",
+ "SaveAudio": "audio", "SaveAudioMP3": "audio", "SaveAudioOpus": "audio"}
+STRING_PRIMITIVES = {"PrimitiveString", "PrimitiveStringMultiline"}
+NUMBER_PRIMITIVES = {"PrimitiveInt": "int", "PrimitiveFloat": "float", "PrimitiveBoolean": "bool"}
+SIZE_OPTIONS = [768, 1024, 1280, 1536]
+# loader class → (models/ folder, phase) when packs/*.json does not know the file
+LOADER_GUESS = [
+ (re.compile(r"Checkpoint", re.I), "checkpoints", "all"),
+ (re.compile(r"Lora", re.I), "loras", "denoise"),
+ (re.compile(r"CLIPVision", re.I), "clip_vision", "encode"),
+ (re.compile(r"CLIP|TextEncoder|T5", re.I), "text_encoders", "encode"),
+ (re.compile(r"VAE", re.I), "vae", "decode"),
+ (re.compile(r"LatentUpscale", re.I), "latent_upscale_models", "aux"),
+ (re.compile(r"Upscale", re.I), "upscale_models", "aux"),
+ (re.compile(r"BackgroundRemoval", re.I), "background_removal", "all"),
+ (re.compile(r"MoGe|Geometry", re.I), "geometry_estimation", "aux"),
+ (re.compile(r"Interpolation|FILM", re.I), "frame_interpolation", "aux"),
+ (re.compile(r"AudioEncoder", re.I), "audio_encoders", "encode"),
+ (re.compile(r"UNET|Diffusion|Model", re.I), "diffusion_models", "denoise"),
+]
+
+
+def is_link(v) -> bool:
+ return isinstance(v, list) and len(v) == 2 and isinstance(v[0], str) and isinstance(v[1], int)
+
+
+def slug(s: str) -> str:
+ s = re.sub(r"[^a-z0-9]+", "_", s.lower()).strip("_")
+ return s if s and s[0].isalpha() else f"in_{s}" if s else "input"
+
+
+def title_of(node: dict) -> str:
+ return ((node.get("_meta") or {}).get("title") or "").strip()
+
+
+def plain_label(title: str, class_type: str) -> str | None:
+ """A title someone gave the node, cleaned up; None when it is just the node's default name."""
+ t = re.sub(r"^(Int|Float|Boolean|String|Primitive\w*)\s*\((.*)\)$", r"\2", title).strip()
+ defaults = {class_type.lower(), re.sub(r"(?<!^)(?=[A-Z])", " ", class_type).lower(), "int", "float", "boolean",
+ "string", "string (multiline)", "primitive", ""}
+ return None if t.lower() in defaults else t
+
+
+# ---------------------------------------------------------------- what Freefield already knows
+def known_files(root: Path) -> tuple[dict[str, dict], dict[str, set[str]], dict[str, str]]:
+ """basename → pack entry, basename → packs holding it, basename → the phase apps use it in."""
+ files: dict[str, dict] = {}
+ in_packs: dict[str, set[str]] = defaultdict(set)
+ for p in sorted((root / "packs").glob("*.json")):
+ if p.name == "index.json":
+ continue
+ spec = json.loads(p.read_text(encoding="utf-8"))
+ for m in spec.get("models", []):
+ files.setdefault(m["name"], m)
+ in_packs[m["name"]].add(spec["pack"])
+ phases: dict[str, Counter] = defaultdict(Counter)
+ for f in sorted((root / "apps").glob("*/app.json")):
+ try:
+ for m in json.loads(f.read_text(encoding="utf-8")).get("models", []):
+ phases[m["file"].rsplit("/", 1)[-1]][m["phase"]] += 1
+ except (OSError, ValueError, KeyError):
+ continue
+ return files, in_packs, {k: c.most_common(1)[0][0] for k, c in phases.items()}
+
+
+# ---------------------------------------------------------------- graph helpers
+def consumers(api: dict) -> dict[str, list[tuple[str, str]]]:
+ """node id → [(consumer node id, consumer input name)]"""
+ out: dict[str, list[tuple[str, str]]] = defaultdict(list)
+ for nid, n in api.items():
+ for k, v in n["inputs"].items():
+ if is_link(v):
+ out[v[0]].append((nid, k))
+ return out
+
+
+def polarity(api: dict, users: dict, start: str) -> str | None:
+ """Follow a text encoder's output downstream until an input named positive / negative takes it."""
+ seen, todo = set(), [start]
+ while todo:
+ nid = todo.pop(0)
+ if nid in seen:
+ continue
+ seen.add(nid)
+ for (c, field) in users.get(nid, []):
+ f = field.lower()
+ if "negative" in f:
+ return "negative"
+ if "positive" in f:
+ return "positive"
+ if f in ("conditioning", "conditioning_to", "conditioning_from", "cond") or "conditioning" in f:
+ todo.append(c)
+ return None
+
+
+def string_source(api: dict, nid: str, field: str) -> tuple[str, str] | None:
+ """Where the words of a text input live: the input itself, or the PrimitiveString node upstream of it."""
+ v = api[nid]["inputs"].get(field)
+ if isinstance(v, str):
+ return (nid, field)
+ seen, todo = set(), [v]
+ while todo:
+ cur = todo.pop(0)
+ if not is_link(cur) or cur[0] in seen or cur[0] not in api:
+ continue
+ seen.add(cur[0])
+ node = api[cur[0]]
+ if node["class_type"] in STRING_PRIMITIVES and isinstance(node["inputs"].get("value"), str):
+ return (cur[0], "value")
+ todo.extend(x for x in node["inputs"].values() if is_link(x))
+ return None
+
+
+def text_fields(node: dict) -> list[str]:
+ ct = node["class_type"]
+ if "Encode" not in ct and "Prompt" not in ct:
+ return []
+ return [k for k in ("text", "prompt", "text_g", "clip_l", "t5xxl") if k in node["inputs"]]
+
+
+# ---------------------------------------------------------------- the import
+def load_workflow(path: Path, object_info: Path | None) -> tuple[dict, list[str]]:
+ data = json.loads(path.read_text(encoding="utf-8"))
+ if isinstance(data, dict) and "nodes" in data and isinstance(data["nodes"], list):
+ info_path = object_info or ROOT / "tools" / "object_info.json"
+ info = json.loads(info_path.read_text(encoding="utf-8"))
+ api, warnings = convert(data, info)
+ return api, [f"converted from the UI format with {info_path.name}"] + warnings
+ if isinstance(data, dict) and isinstance(data.get("prompt"), dict): # a saved /prompt request
+ data = data["prompt"]
+ if not isinstance(data, dict) or not all(isinstance(n, dict) and "class_type" in n for n in data.values()):
+ raise ValueError("not a ComfyUI workflow: expected API format {id: {class_type, inputs}} or a UI workflow")
+ return data, []
+
+
+def build(api: dict, app_id: str, name: str | None, section: str | None, source: str, root: Path = ROOT,
+ object_info: dict | None = None) -> tuple[dict, dict, dict]:
+ """(app.json, workflow.api.json, report) for a workflow already in API format."""
+ api = json.loads(json.dumps(api))
+ info = object_info or {}
+ report: dict = {"inputs": [], "outputs": [], "models": [], "unknown": [], "check": [], "pack": None}
+ files, in_packs, phases = known_files(root)
+ users = consumers(api)
+ order = sorted(api, key=lambda k: [int(p) if p.isdigit() else p for p in k.split(":")])
+ inputs: list[dict] = []
+ taken: Counter = Counter()
+
+ def add(inp: dict) -> None:
+ base = inp["id"]
+ taken[base] += 1
+ if taken[base] > 1:
+ inp["id"] = f"{base}{taken[base]}"
+ inputs.append(inp)
+
+ # files the user gives
+ for nid in order:
+ n = api[nid]
+ if n["class_type"] in FILE_LOADERS:
+ kind, field = FILE_LOADERS[n["class_type"]]
+ label = plain_label(title_of(n), n["class_type"]) or {"image": "Image", "video": "Video", "audio": "Audio"}[kind]
+ if label.lower().startswith("load "):
+ label = label[5:].strip().capitalize() or label
+ add({"id": kind, "type": kind, "label": label, "required": True, "bind": [{"node": nid, "field": field}]})
+
+ # prompts
+ seen_src: set[tuple[str, str]] = set()
+ for nid in order:
+ n = api[nid]
+ for field in text_fields(n):
+ src = string_source(api, nid, field)
+ if not src or src in seen_src:
+ continue
+ words = api[src[0]]["inputs"][src[1]]
+ pol = polarity(api, users, nid) or ("positive" if n["class_type"].startswith("TextGenerate") or field == "prompt" else None)
+ if pol is None:
+ continue
+ if pol == "negative" and not words.strip():
+ continue
+ seen_src.add(src)
+ if pol == "positive":
+ add({"id": "prompt", "type": "textarea", "label": "Prompt", "required": True, "default": words,
+ "bind": [{"node": src[0], "field": src[1]}]})
+ else:
+ add({"id": "negative", "type": "textarea", "label": "Negative prompt", "default": words,
+ "bind": [{"node": src[0], "field": src[1]}]})
+
+ # numbers on samplers and latents
+ seeds, steps = [], []
+ for nid in order:
+ n = api[nid]
+ for k, v in n["inputs"].items():
+ if "." in k or isinstance(v, bool) or not isinstance(v, int):
+ continue
+ if k in ("seed", "noise_seed") and ("Sampler" in n["class_type"] or "Noise" in n["class_type"]):
+ seeds.append({"node": nid, "field": k})
+ elif k == "steps" and ("Sampler" in n["class_type"] or "Scheduler" in n["class_type"]):
+ steps.append((nid, v))
+ for i, (nid, v) in enumerate(steps):
+ add({"id": "steps", "type": "slider", "label": "Steps" if len(steps) == 1 else f"Steps ({i + 1})",
+ "min": max(1, v // 2), "max": max(v * 2, 4), "step": 1, "default": v, "bind": [{"node": nid, "field": "steps"}]})
+ for nid in order:
+ n = api[nid]
+ if "Latent" not in n["class_type"] or not n["class_type"].startswith("Empty"):
+ continue
+ for k in ("width", "height"):
+ v = n["inputs"].get(k)
+ if isinstance(v, int) and not isinstance(v, bool):
+ opts = sorted(set(SIZE_OPTIONS) | {v})
+ add({"id": k, "type": "select", "label": k.capitalize(), "default": v,
+ "options": [{"value": o, "label": str(o)} for o in opts], "bind": [{"node": nid, "field": k}]})
+ if n["class_type"].startswith("Empty") and isinstance(n["inputs"].get("length"), int):
+ v = n["inputs"]["length"]
+ add({"id": "frames", "type": "slider", "label": "Frames", "min": 1, "max": max(v * 2, 2), "step": 1,
+ "default": v, "bind": [{"node": nid, "field": "length"}]})
+ for nid in order:
+ n = api[nid]
+ if n["class_type"] == "ResolutionSelector" and isinstance(n["inputs"].get("aspect_ratio"), str):
+ spec = ((info.get("ResolutionSelector") or {}).get("input", {}).get("required", {}) or {}).get("aspect_ratio")
+ choices = spec[0] if spec and isinstance(spec[0], list) else [n["inputs"]["aspect_ratio"]]
+ add({"id": "aspect", "type": "select", "label": "Aspect ratio", "default": n["inputs"]["aspect_ratio"],
+ "options": [{"value": c, "label": c} for c in choices], "bind": [{"node": nid, "field": "aspect_ratio"}]})
+ # primitives somebody named
+ for nid in order:
+ n = api[nid]
+ kind = NUMBER_PRIMITIVES.get(n["class_type"])
+ v = n["inputs"].get("value")
+ label = plain_label(title_of(n), n["class_type"])
+ if not kind or label is None or is_link(v):
+ continue
+ if kind == "bool" and isinstance(v, bool):
+ add({"id": slug(label), "type": "toggle", "label": label, "default": v, "bind": [{"node": nid, "field": "value"}]})
+ elif kind in ("int", "float") and isinstance(v, (int, float)) and not isinstance(v, bool):
+ hi = max(v * 2, 1) if v > 0 else 10
+ add({"id": slug(label), "type": "slider", "label": label, "min": 0 if v <= 0 else (1 if kind == "int" else 0),
+ "max": hi, "step": 1 if kind == "int" else 0.05, "default": v, "bind": [{"node": nid, "field": "value"}]})
+ report["check"].append(f'"{label}" ({nid}) became a slider {0 if v <= 0 else 1}–{hi}: keep it only if a user should change it')
+ if seeds:
+ add({"id": "seed", "type": "seed", "label": "Seed", "default": -1, "help": "-1 picks a random seed", "bind": seeds})
+ if len(seeds) > 1:
+ report["check"].append(f"the seed writes {len(seeds)} nodes ({', '.join(s['node'] for s in seeds)}): "
+ "keep only the ones that should change between runs")
+
+ # outputs
+ outputs = []
+ for nid in order:
+ n = api[nid]
+ if n["class_type"] == "PreviewImage":
+ n["class_type"] = "SaveImage"
+ n["_meta"] = {"title": "Save Image"}
+ report["check"].append(f"PreviewImage {nid} became a SaveImage")
+ if n["class_type"] in SAVE_TYPES:
+ n["inputs"]["filename_prefix"] = f"freefield/{app_id}"
+ t = SAVE_TYPES[n["class_type"]]
+ outputs.append({"node": nid, "type": t, "label": t.capitalize()})
+ if len(outputs) > 1:
+ report["check"].append(f"{len(outputs)} outputs: the first one listed is the app's main result")
+
+ # models
+ models: list[dict] = []
+ seen_files: set[str] = set()
+ for nid in order:
+ n = api[nid]
+ for k, v in n["inputs"].items():
+ if not isinstance(v, str) or not v.lower().endswith(MODEL_EXT):
+ continue
+ base = v.replace("\\", "/").rsplit("/", 1)[-1]
+ if base in seen_files:
+ continue
+ seen_files.add(base)
+ guess_dir, guess_phase = next(((d, p) for rx, d, p in LOADER_GUESS if rx.search(n["class_type"])),
+ ("diffusion_models", "denoise"))
+ known = files.get(base)
+ if known:
+ m = {"repo": known["repo"], "file": known["file"], "dir": known["dir"], "gb": known["gb"],
+ "phase": phases.get(base, guess_phase)}
+ if known.get("gated"):
+ m["gated"] = True
+ if known.get("license"):
+ m["license"] = known["license"]
+ models.append(m)
+ report["models"].append(f"{base} {known['repo']} {known['gb']} GB ({m['phase']})")
+ else:
+ models.append({"repo": "TODO/unknown", "file": base, "dir": guess_dir, "gb": 0.001, "phase": guess_phase})
+ report["unknown"].append(f"{base} (loaded by {n['class_type']} {nid}.{k}; folder guessed as {guess_dir})")
+
+ # pack
+ known_bases = [m["file"].rsplit("/", 1)[-1] for m in models if m["repo"] != "TODO/unknown"]
+ pack = app_id
+ if known_bases and not report["unknown"]:
+ common = set.intersection(*(in_packs[b] for b in known_bases))
+ if common:
+ sizes = {p: sum(1 for s in in_packs.values() if p in s) for p in common}
+ pack = min(common, key=lambda p: (sizes[p], p))
+ report["pack"] = pack if pack != app_id else f"{app_id} (new: no existing pack holds every model)"
+
+ if not section:
+ kinds = {o["type"] for o in outputs}
+ section = "video" if "video" in kinds else "audio" if kinds == {"audio"} else "image"
+ notes = ["Imported by tools/import_workflow.py: check every input and model before marking it ready."]
+ if report["unknown"]:
+ notes.append("Models with repo TODO/unknown: find their Hugging Face repo, path and size, then run "
+ "python tools/catalog.py all.")
+ app = {
+ "$schema": "../../schema/app.schema.json",
+ "id": app_id,
+ "name": (name or app_id.replace("-", " ").title())[:40],
+ "section": section,
+ "pack": pack,
+ "status": "draft",
+ "summary": f"Imported from {source}."[:160],
+ "workflow": "workflow.api.json",
+ "source": {"template": source, "notes": " ".join(notes)},
+ "inputs": inputs,
+ "outputs": outputs,
+ "models": models,
+ }
+ report["inputs"] = [f"{i['id']} ({i['type']}) → " + ", ".join(f"{b['node']}.{b['field']}" for b in i["bind"])
+ for i in inputs]
+ report["outputs"] = [f"{o['type']} ← {o['node']} ({api[o['node']]['class_type']})" for o in outputs]
+ if not outputs:
+ report["check"].append("no Save node: the app needs at least one output")
+ return app, api, report
+
+
+def validate(app: dict, api: dict, root: Path = ROOT) -> tuple[list[str], list[str]] | None:
+ """Run the repo's validator on this app alone, in a scratch copy of the repo. (errors, warnings), or None."""
+ node = shutil.which("node")
+ if not node or not (root / "node_modules").exists():
+ return None
+ with tempfile.TemporaryDirectory() as tmp:
+ t = Path(tmp)
+ shutil.copytree(root / "server", t / "server")
+ shutil.copytree(root / "schema", t / "schema")
+ shutil.copy(root / "package.json", t / "package.json")
+ (t / "node_modules").symlink_to(root / "node_modules", target_is_directory=True)
+ d = t / "apps" / app["id"]
+ d.mkdir(parents=True)
+ (d / "app.json").write_text(json.dumps(app, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+ (d / "workflow.api.json").write_text(json.dumps(api, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+ r = subprocess.run([node, "--experimental-strip-types", "--disable-warning=ExperimentalWarning",
+ "server/cli.ts", "validate"], cwd=t, capture_output=True, text=True, encoding="utf-8")
+ lines = r.stdout.splitlines()
+ errors = [l[2:] for l in lines if l.startswith("✗ ")]
+ warnings = [l[2:] for l in lines if l.startswith("! ")]
+ if r.returncode != 0 and not errors:
+ errors = [(r.stderr or r.stdout).strip() or f"validator exited {r.returncode}"]
+ return errors, warnings
+
+
+def print_report(app: dict, report: dict, checked, out: Path) -> None:
+ print(f"{app['id']}: {len(app['inputs'])} inputs, {len(app['outputs'])} outputs, {len(app['models'])} models → {out}")
+ for title, key in (("inputs", "inputs"), ("outputs", "outputs"), ("models matched in packs/", "models"),
+ ("models Freefield does not know", "unknown"), ("check by hand", "check")):
+ if report[key]:
+ print(f" {title}:")
+ for line in report[key]:
+ print(f" - {line}")
+ print(f" pack: {report['pack']}")
+ if checked is None:
+ print(" validator: not run (node or node_modules missing)")
+ else:
+ errors, warnings = checked
+ print(f" validator: {len(errors)} errors, {len(warnings)} warnings")
+ for line in errors + warnings:
+ print(f" {line}")
+
+
+def main(argv: list[str] | None = None) -> int:
+ ap = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+ ap.add_argument("workflow", type=Path)
+ ap.add_argument("--id", required=True, help="app id: lowercase words joined by - (the folder name)")
+ ap.add_argument("--name", help="name shown in the UI (default: the id in title case)")
+ ap.add_argument("--section", choices=["image", "video", "audio", "utility"], help="default: from the outputs")
+ ap.add_argument("--out", type=Path, help="folder to write (default: apps/<id>)")
+ ap.add_argument("--object-info", type=Path, help="GET /object_info of a ComfyUI, for UI workflows and option lists")
+ ap.add_argument("--force", action="store_true", help="overwrite the folder when it exists")
+ a = ap.parse_args(argv)
+ if not re.fullmatch(r"[a-z0-9]+(-[a-z0-9]+)*", a.id):
+ print("--id must be lowercase letters and digits joined by -", file=sys.stderr)
+ return 2
+ out = a.out or ROOT / "apps" / a.id
+ if (out / "app.json").exists() and not a.force:
+ print(f"{out / 'app.json'} exists; pass --force to overwrite", file=sys.stderr)
+ return 2
+ try:
+ api, notes = load_workflow(a.workflow, a.object_info)
+ except (ValueError, ConversionError, OSError) as e:
+ print(f"{a.workflow}: {e}", file=sys.stderr)
+ return 1
+ info_path = a.object_info or ROOT / "tools" / "object_info.json"
+ info = json.loads(info_path.read_text(encoding="utf-8")) if info_path.exists() else {}
+ app, api, report = build(api, a.id, a.name, a.section, a.workflow.name, object_info=info)
+ report["check"] = notes + report["check"]
+ checked = validate(app, api)
+ out.mkdir(parents=True, exist_ok=True)
+ (out / "app.json").write_text(json.dumps(app, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+ (out / "workflow.api.json").write_text(json.dumps(api, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+ print_report(app, report, checked, out)
+ if not app["outputs"] or (checked is not None and checked[0]):
+ return 1
+ return 0
+
+
+if __name__ == "__main__":
+ sys.exit(main())
diff --git a/tools/object_info.json b/tools/object_info.json
index 7b0804b..383e7b8 100644
--- a/tools/object_info.json
+++ b/tools/object_info.json
@@ -1 +1 @@
-{"AudioMerge":{"input":{"required":{"audio1":["AUDIO",{}],"audio2":["AUDIO",{}],"merge_method":["COMBO",{"tooltip":"The method used to combine the audio waveforms.","multiselect":false,"options":["add","mean","subtract","multiply"]}]}},"input_order":{"required":["audio1","audio2","merge_method"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"AudioMerge","display_name":"Merge Audio","description":"Combine two audio tracks by overlaying their waveforms.","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["mix audio","overlay audio","layer audio"],"essentials_category":null,"has_intermediate_output":false},"BasicGuider":{"input":{"required":{"model":["MODEL",{}],"conditioning":["CONDITIONING",{}]}},"input_order":{"required":["model","conditioning"]},"is_input_list":false,"output":["GUIDER"],"output_is_list":[false],"output_name":["GUIDER"],"output_tooltips":[null],"output_matchtypes":null,"name":"BasicGuider","display_name":"Basic Guider","description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/guiders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"BasicScheduler":{"input":{"required":{"model":["MODEL",{}],"scheduler":["COMBO",{"multiselect":false,"options":["simple","sgm_uniform","karras","exponential","ddim_uniform","beta","normal","linear_quadratic","kl_optimal"]}],"steps":["INT",{"default":20,"min":1,"max":10000}],"denoise":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}]}},"input_order":{"required":["model","scheduler","steps","denoise"]},"is_input_list":false,"output":["SIGMAS"],"output_is_list":[false],"output_name":["SIGMAS"],"output_tooltips":[null],"output_matchtypes":null,"name":"BasicScheduler","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/schedulers","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"BatchImagesNode":{"input":{"required":{"images":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"image":["IMAGE",{}]}},"prefix":"image","min":1,"max":50}}]}},"input_order":{"required":["images"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"BatchImagesNode","display_name":"Batch Images","description":"","python_module":"comfy_extras.nodes_post_processing","category":"image/batch","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["batch","image batch","batch images","combine images","merge images","stack images"],"essentials_category":"Image Tools","has_intermediate_output":false},"CFGGuider":{"input":{"required":{"model":["MODEL",{}],"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}]}},"input_order":{"required":["model","positive","negative","cfg"]},"is_input_list":false,"output":["GUIDER"],"output_is_list":[false],"output_name":["GUIDER"],"output_tooltips":[null],"output_matchtypes":null,"name":"CFGGuider","display_name":"CFG Guider","description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/guiders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CFGNorm":{"input":{"required":{"model":["MODEL",{}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":100.0,"step":0.01}]},"optional":{"pre_cfg":["BOOLEAN",{"tooltip":"If true, rescale the combined noise BEFORE the sampler's CFG combine, without clamping (can amplify). Matches the norm-scaled CFG used by models like Lens. Default false keeps the original post-CFG x0-space attenuate-only behavior.","default":false}]}},"input_order":{"required":["model","strength"],"optional":["pre_cfg"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["patched_model"],"output_tooltips":[null],"output_matchtypes":null,"name":"CFGNorm","display_name":null,"description":"","python_module":"comfy_extras.nodes_cfg","category":"advanced/guidance","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CLIPLoader":{"input":{"required":{"clip_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]],"type":[["stable_diffusion","stable_cascade","sd3","stable_audio","mochi","ltxv","pixart","cosmos","lumina2","wan","hidream","chroma","ace","omnigen2","qwen_image","hunyuan_image","flux2","ovis","longcat_image","cogvideox","lens","pixeldit","ideogram4","boogu","krea2","joyimage","mage","minimax","yue2","flux3","moss_image"]]},"optional":{"device":[["default","cpu"],{"advanced":true}]}},"input_order":{"required":["clip_name","type"],"optional":["device"]},"is_input_list":false,"output":["CLIP"],"output_is_list":[false],"output_name":["CLIP"],"name":"CLIPLoader","display_name":"Load CLIP","description":"Recipes:\nsd: clip-l\nstable cascade: clip-g\nsd3: t5 xxl / clip-g / clip-l\nstable audio: t5 base\nmochi: t5 xxl\ncogvideox: t5 xxl (226-token padding)\ncosmos: old t5 xxl\nlumina2: gemma 2 2B\nwan: umt5 xxl\nhidream: llama-3.1 (Recommend) or t5\nomnigen2: qwen vl 2.5 3B\njoyimage: qwen3-vl 8B\nlens: gpt-oss-20b\npixeldit: gemma 2 2B elm\nminimax: MiniMax H3 Qwen3-VL or Music3 Qwen/RVQ\nflux3: qwen3-vl 4B","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"CLIPTextEncode":{"input":{"required":{"text":["STRING",{"multiline":true,"dynamicPrompts":true,"tooltip":"The text to be encoded."}],"clip":["CLIP",{"tooltip":"The CLIP model used for encoding the text."}]}},"input_order":{"required":["text","clip"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"name":"CLIPTextEncode","display_name":"CLIP Text Encode (Prompt)","description":"Encodes a text prompt using a CLIP model into an embedding that can be used to guide the diffusion model towards generating specific images.","python_module":"nodes","category":"model/conditioning","output_node":false,"has_intermediate_output":false,"output_tooltips":["A conditioning containing the embedded text used to guide the diffusion model."],"search_aliases":["text","prompt","text prompt","positive prompt","negative prompt","encode text","text encoder","encode prompt"]},"CLIPVisionEncode":{"input":{"required":{"clip_vision":["CLIP_VISION"],"image":["IMAGE"],"crop":[["center","none"]]}},"input_order":{"required":["clip_vision","image","crop"]},"is_input_list":false,"output":["CLIP_VISION_OUTPUT"],"output_is_list":[false],"output_name":["CLIP_VISION_OUTPUT"],"name":"CLIPVisionEncode","display_name":"CLIP Vision Encode","description":"","python_module":"nodes","category":"model/conditioning","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"CLIPVisionLoader":{"input":{"required":{"clip_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]]}},"input_order":{"required":["clip_name"]},"is_input_list":false,"output":["CLIP_VISION"],"output_is_list":[false],"output_name":["CLIP_VISION"],"name":"CLIPVisionLoader","display_name":"Load CLIP Vision","description":"","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"Canny":{"input":{"required":{"image":["IMAGE",{}],"low_threshold":["FLOAT",{"default":0.4,"min":0.01,"max":0.99,"step":0.01}],"high_threshold":["FLOAT",{"default":0.8,"min":0.01,"max":0.99,"step":0.01}]}},"input_order":{"required":["image","low_threshold","high_threshold"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"Canny","display_name":"Detect Edges (Canny)","description":"","python_module":"comfy_extras.nodes_canny","category":"image/filters","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["edge detection","outline","contour detection","line art"],"essentials_category":"Image Tools","has_intermediate_output":false},"CheckpointLoaderSimple":{"input":{"required":{"ckpt_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"],{"tooltip":"The name of the checkpoint (model) to load."}]}},"input_order":{"required":["ckpt_name"]},"is_input_list":false,"output":["MODEL","CLIP","VAE"],"output_is_list":[false,false,false],"output_name":["MODEL","CLIP","VAE"],"name":"CheckpointLoaderSimple","display_name":"Load Checkpoint","description":"Loads a diffusion model checkpoint, diffusion models are used to denoise latents.","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"output_tooltips":["The model used for denoising latents.","The CLIP model used for encoding text prompts.","The VAE model used for encoding and decoding images to and from latent space."],"search_aliases":["load model","checkpoint","model loader","load checkpoint","ckpt","model"]},"ColorTransfer":{"input":{"required":{"image_target":["IMAGE",{"tooltip":"Image(s) to apply the color transform to."}],"image_ref":["IMAGE",{"tooltip":"Reference image(s) to match colors to."}],"method":["COMBO",{"multiselect":false,"options":["reinhard_lab","mkl_lab","histogram"]}],"source_stats":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"per_frame: each frame matched to image_ref individually. uniform: pool stats across all source frames as baseline, match to image_ref. target_frame: use one chosen frame as the baseline for the transform to image_ref, applied uniformly to all frames (preserves relative differences)","options":[{"key":"per_frame","inputs":{"required":{}}},{"key":"uniform","inputs":{"required":{}}},{"key":"target_frame","inputs":{"required":{"target_index":["INT",{"tooltip":"Frame index used as the source baseline for computing the transform to image_ref","default":0,"min":0,"max":10000}]}}}]}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":10.0,"step":0.01}]}},"input_order":{"required":["image_target","image_ref","method","source_stats","strength"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["image"],"output_tooltips":[null],"output_matchtypes":null,"name":"ColorTransfer","display_name":"Transfer Color","description":"Match the colors of one image to another using various algorithms.","python_module":"comfy_extras.nodes_post_processing","category":"image/filters","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["color match","color grading","color correction","match colors","color transform","mkl","reinhard","histogram"],"essentials_category":null,"has_intermediate_output":false},"ComfyMathExpression":{"input":{"required":{"expression":["STRING",{"default":"a + b","multiline":true}],"values":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"value":["FLOAT,INT,BOOLEAN",{}]}},"names":["a","b","c","d","e","f","g","h","i","j","k","l","m","n","o","p","q","r","s","t","u","v","w","x","y","z"],"min":1}}]}},"input_order":{"required":["expression","values"]},"is_input_list":false,"output":["FLOAT","INT","BOOLEAN"],"output_is_list":[false,false,false],"output_name":["FLOAT","INT","BOOL"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"ComfyMathExpression","display_name":"Math Expression","description":"","python_module":"comfy_extras.nodes_math","category":"utilities","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["math","expression","formula","calculate","calculator","eval"],"essentials_category":null,"has_intermediate_output":false},"ComfyNotNode":{"input":{"required":{"value":["*",{}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["BOOLEAN"],"output_is_list":[false],"output_name":["BOOLEAN"],"output_tooltips":[null],"output_matchtypes":null,"name":"ComfyNotNode","display_name":"Not","description":"Logical NOT operation. Returns true if the value is falsy. Uses Python's rules for truthiness.","python_module":"comfy_extras.nodes_logic","category":"utilities/logic","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["invert","toggle","negate","flip boolean"],"essentials_category":null,"has_intermediate_output":false},"ComfySwitchNode":{"input":{"required":{"switch":["BOOLEAN",{}]},"optional":{"on_false":["COMFY_MATCHTYPE_V3",{"lazy":true,"template":{"template_id":"switch","allowed_types":"*"}}],"on_true":["COMFY_MATCHTYPE_V3",{"lazy":true,"template":{"template_id":"switch","allowed_types":"*"}}]}},"input_order":{"required":["switch"],"optional":["on_false","on_true"]},"is_input_list":false,"output":["COMFY_MATCHTYPE_V3"],"output_is_list":[false],"output_name":["output"],"output_tooltips":[null],"output_matchtypes":["switch"],"name":"ComfySwitchNode","display_name":"If/Else Switch","description":"","python_module":"comfy_extras.nodes_logic","category":"utilities/logic","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["if","then","switch","conditional","branch"],"essentials_category":null,"has_intermediate_output":false},"ConditioningZeroOut":{"input":{"required":{"conditioning":["CONDITIONING"]}},"input_order":{"required":["conditioning"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"name":"ConditioningZeroOut","display_name":"Conditioning Zero Out","description":"","python_module":"nodes","category":"model/conditioning/transform","output_node":false,"has_intermediate_output":false,"search_aliases":["null conditioning","clear conditioning"]},"ContextWindowsManual":{"input":{"required":{"model":["MODEL",{"tooltip":"The model to apply context windows to during sampling."}],"context_length":["INT",{"tooltip":"The length of the context window.","default":16,"min":1}],"context_overlap":["INT",{"tooltip":"The overlap of the context window.","default":4,"min":0}],"context_schedule":["COMBO",{"tooltip":"Step-dependent scheduling algorithm for context windows.","default":"standard_static","multiselect":false,"options":["standard_static","standard_uniform","looped_uniform","batched"]}],"context_stride":["INT",{"tooltip":"The stride of the context window; only applicable to uniform schedules.","default":1,"min":1}],"closed_loop":["BOOLEAN",{"tooltip":"Whether to close the context window loop; only applicable to looped schedules.","default":false}],"fuse_method":["COMBO",{"tooltip":"The method to use to fuse the context windows.","default":"pyramid","multiselect":false,"options":["pyramid","relative","flat","overlap-linear"]}],"dim":["INT",{"tooltip":"The dimension to apply the context windows to.","default":0,"min":0,"max":5}],"freenoise":["BOOLEAN",{"tooltip":"Whether to apply FreeNoise noise shuffling, improves window blending.","default":false}],"cond_retain_index_list":["STRING",{"tooltip":"List of latent indices to retain in the conditioning tensors for each window. For concat-style I2V models (e.g. Wan I2V, HunyuanVideo I2V, Cosmos I2V, SVD) the encoded start image lives in the c_concat conditioning channels; setting this to '0' will retain that start image content at sub-pos 0 of every window.","default":"","multiline":false}],"split_conds_to_windows":["BOOLEAN",{"tooltip":"Whether to split multiple conditionings (created by ConditionCombine) to each window based on region index.","default":false}],"latent_retain_index_list":["STRING",{"tooltip":"List of latent indices to retain in the noise latent itself for each window. Use for workflows where reference content (e.g. a start image) lives directly in the noise latent rather than in separate conditioning channels (e.g. inplace-style I2V like LTXV, AnimateDiff). Independent of cond_retain_index_list.","default":"","multiline":false}],"causal_window_fix":["BOOLEAN",{"tooltip":"Whether to add a causal fix frame to non-0-indexed context windows.","default":true}]}},"input_order":{"required":["model","context_length","context_overlap","context_schedule","context_stride","closed_loop","fuse_method","dim","freenoise","cond_retain_index_list","split_conds_to_windows","latent_retain_index_list","causal_window_fix"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"output_tooltips":["The model with context windows applied during sampling."],"output_matchtypes":null,"name":"ContextWindowsManual","display_name":"Context Windows (Manual)","description":"Manually set context windows.","python_module":"comfy_extras.nodes_context_windows","category":"model/patch","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CreateBoundingBoxes":{"input":{"required":{"width":["INT",{"tooltip":"Width of the canvas and the pixel grid for the bounding boxes.","default":1024,"min":64,"max":16384,"step":16}],"height":["INT",{"tooltip":"Height of the canvas and the pixel grid for the bounding boxes.","default":1024,"min":64,"max":16384,"step":16}],"editor_state":["BOUNDING_BOXES",{"tooltip":"Draw bounding boxes and set each box type, text, description, color palette. Start with background element first and foreground last.","default":[],"socketless":false}]},"optional":{"background":["IMAGE",{"tooltip":"Optional image used as background in the canvas and preview."}],"bboxes":["BOUNDING_BOX,ARRAY,STRING",{"tooltip":"Bounding boxes, elements, or a JSON string to initialize the canvas. A new upstream value initializes the canvas; edits made on the canvas take priority and are kept until the upstream value changes again."}],"last_incoming":["BOUNDING_BOXES",{"tooltip":"Internal state managed by the canvas: the upstream bboxes value that last initialized it. Leave empty to re-initialize the canvas from the bboxes input on the next run.","default":[],"socketless":true}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["width","height","editor_state"],"optional":["background","bboxes","last_incoming"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["IMAGE","BOUNDING_BOX","ARRAY"],"output_is_list":[false,false,false],"output_name":["preview","bboxes","elements"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"CreateBoundingBoxes","display_name":"Create Bounding Boxes","description":"Draw bounding boxes in a canvas. Outputs Ideogram prompt elements, pixel-space bounding boxes, and a preview image.","python_module":"comfy_extras.nodes_bounding_boxes","category":"utilities","output_node":true,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CreateVideo":{"input":{"required":{"images":["IMAGE",{"tooltip":"The images to create a video from."}],"fps":["FLOAT",{"default":30.0,"min":1.0,"max":120.0,"step":1.0}]},"optional":{"audio":["AUDIO",{"tooltip":"The audio to add to the video."}],"bit_depth":["COMBO",{"tooltip":"Auto uses 8-bit for sRGB and 10-bit for HDR. Explicit 8-bit and 10-bit choices are independent of colorspace.","default":"auto","multiselect":false,"options":["auto",8,10]}],"color_space":["COMBO",{"tooltip":"Colorspace of the input images. HDR selects BT.2020/HLG and HDR PQ selects BT.2020/PQ.","default":"sRGB","multiselect":false,"options":["sRGB","HDR","HDR PQ"]}],"codec":["COMBO",{"tooltip":"Optionally encode the video immediately. None keeps the images in tensor form; Auto uses H.264.","advanced":true,"default":"none","multiselect":false,"options":["none","auto","h264","av1"]}]}},"input_order":{"required":["images","fps"],"optional":["audio","bit_depth","color_space","codec"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["VIDEO"],"output_tooltips":[null],"output_matchtypes":null,"name":"CreateVideo","display_name":"Create Video","description":"Create a video from images.","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["images to video"],"essentials_category":"Video Tools","has_intermediate_output":false},"CrossViewPrepareClip":{"input":{"required":{"video":["VIDEO"],"duration":["FLOAT",{"default":5.0,"min":0.2,"max":15.0,"step":0.1,"tooltip":"Seconds of the clip to use, from its start. Snapped to the nearest frame count H3 accepts (17k+5 at 24 fps) and capped at the clip."}],"aspect_ratio":[["source","16:9","9:16","1:1","4:3","3:4","21:9"],{"default":"source","tooltip":"Output shape. 'source' keeps the clip's own ratio; the others centre-crop to it."}],"megapixels":["FLOAT",{"default":0.5,"min":0.1,"max":4.0,"step":0.05}],"multiple":["INT",{"default":32,"min":8,"max":128,"step":8}]}},"input_order":{"required":["video","duration","aspect_ratio","megapixels","multiple"]},"is_input_list":false,"output":["IMAGE","INT","INT","INT","AUDIO"],"output_is_list":[false,false,false,false,false],"output_name":["frames","width","height","length","audio"],"name":"CrossViewPrepareClip","display_name":"CrossView Prepare Clip","description":"","python_module":"custom_nodes.ComfyUI-CrossViewWarp","category":"CrossView","output_node":false,"has_intermediate_output":false,"output_tooltips":["The frames both stages use, at 24 fps.","Frame width.","Frame height.","Frame count, on H3's 17k+5 grid.","The clip's own audio over exactly those frames, silence if it has none."],"search_aliases":[]},"CrossViewWarp":{"input":{"required":{"frames":["IMAGE",{"tooltip":"Input video frames. Feed the same frames you send to the depth node."}],"azimuth":["FLOAT",{"default":-30.0,"min":-180.0,"max":180.0,"step":1.0,"tooltip":"Horizontal orbit angle (deg). Negative orbits LEFT, positive RIGHT. The strongest control: +-45 is where the model has been checked, and the training set covers out to +-90 evenly. Past 90 the hidden side is invented."}],"elevation":["FLOAT",{"default":20.0,"min":-90.0,"max":90.0,"step":1.0,"tooltip":"Vertical orbit angle (deg). Positive rises above the subject, negative looks up from below. UP is the weak direction: only 3.1% of the training set sits below -20."}],"distance":["FLOAT",{"default":1.0,"min":0.1,"max":3.0,"step":0.05,"tooltip":"Camera distance (1.0 = same as the source). Below 1 moves closer, above 1 pulls back - and pulling back reveals area the source never framed, so the magenta holes grow. The current LoRA does not follow this reliably."}],"hfov":["FLOAT",{"default":50.0,"min":0.0,"max":120.0,"step":1.0,"tooltip":"Assumed horizontal field of view (deg). ~50 is a normal lens; change it if the clip is clearly wide-angle or telephoto. 0 takes it from moge_geometry, which runs about 10% short - set it yourself if you know the lens."}],"vertical_shift":["FLOAT",{"default":0.0,"min":-1.0,"max":1.0,"step":0.02,"tooltip":"Vertical lens shift, as a fraction of image height. Positive moves the picture DOWN, i.e. the framing up, which rescues a clipped head. The camera does not move, so parallax is unchanged. Leave at 0 unless the framing needs it."}],"depth_ratio":["FLOAT",{"default":6.0,"min":1.5,"max":1000.0,"step":0.5,"tooltip":"Max far/near depth ratio of the scene. Lower = flatter relief and a cleaner warp; higher = more parallax but messier on cluttered scenes. Close-up faces: 2.5-4; mid shots: 4-8; deep/wide scenes: 8-16."}],"smooth_depth":["BOOLEAN",{"default":false,"tooltip":"Edge-aware depth smoothing before warping. Turn ON if the warp has too many speckle holes; it trades a little sharpness for cleaner disocclusion."}],"invert_depth":["BOOLEAN",{"default":false,"tooltip":"Flip depth polarity (near<->far). Leave OFF for the standard DA-V2 node. Turn ON only if the warp looks inside-out - background moving like foreground, or the subject caving in."}]},"optional":{"roll_lock":["BOOLEAN",{"default":true,"tooltip":"Retired and ignored. It rolled the camera to hold the subject's projected lean, but the orbit introduces no roll to correct - a level source gives a level target at every angle - and what it was reacting to is keystone from elevation, which a roll cannot fix. Both training datasets have exactly 0 camera roll."}],"pivot_override":["BOOLEAN",{"default":true,"tooltip":"Orbit around the explicit pivot below instead of an auto- estimated one. ON by default, which is the recommended setup."}],"pivot_x":["FLOAT",{"default":0.0,"min":-1000.0,"max":1000.0,"step":0.01,"tooltip":"Pivot X in the source camera frame (+ = right), in depth units. 0 = image centre. The camera aims at the pivot, so moving this also reframes the shot - unless keep_source_aim is ON, where it only moves the orbit centre."}],"pivot_y":["FLOAT",{"default":0.0,"min":-1000.0,"max":1000.0,"step":0.01,"tooltip":"Pivot Y (+ = down), in depth units. 0 = image centre. Doubles as a reframe like pivot_x. Use vertical_shift to reframe without moving the camera."}],"pivot_z":["FLOAT",{"default":1.05,"min":0.01,"max":1000.0,"step":0.01,"tooltip":"Pivot depth. ~1.05 sits on the nearest subject in the middle of the frame; raise it to orbit something further back. The orbit radius is distance * |pivot|, so this widens the arc as well as moving its centre."}],"use_keyframes":["BOOLEAN",{"default":false,"tooltip":"Animate the camera along the keyframes path instead of holding one pose. Switched on for you when the first keyframe is placed."}],"frame_count":["INT",{"default":0,"min":0,"max":100000,"step":1,"tooltip":"Retired and ignored. The clip length comes from the cached preview."}],"keyframes":["STRING",{"default":"","multiline":false,"tooltip":"Camera path as JSON - normally written for you by the KEY button on the preview, or by right-clicking the orbit sphere. Each entry: 'f' is the frame number counted from 1, 'az'/'el' are degrees, 'dist' is the distance. Optional 'vs' (vertical shift) and 'px'/'py'/'pz' (the pivot) inherit the static widgets when left out; px/py/pz need pivot_override ON. Example: [{\"f\":1,\"az\":-30,\"el\":20,\"dist\":1.0},{\"f\":49,\"az\":45,\"el\":10,\"dist\":1.2}] . The pose is held before the first and after the last keyframe, so a path may cover only part of the clip. Empty = no move."}],"interp_motion":[["linear","ease_in","ease_out","ease_in_out","smooth"],{"default":"linear","tooltip":"Shape and timing of the camera path. linear = straight legs at constant speed. ease_in / ease_out / ease_in_out = the same legs with the camera settling into every keyframe. smooth = a Catmull-Rom spline gliding through them with no corner. Only used while use_keyframes is ON."}],"interpolation":[["linear","smooth"],{"default":"linear","tooltip":"Driven by interp_motion."}],"keep_source_aim":["BOOLEAN",{"default":false,"tooltip":"Which point the orbiting camera looks at. OFF: it looks at the pivot, so the pivot sits dead centre and moving it reframes the shot. ON: it looks where the SOURCE camera looked, so your original framing is kept and the pivot is only the orbit centre. Needs pivot_x or pivot_y off zero to do anything - on the optical axis both aims are the same point. ON is what the training data does and scores 28.9 against 42.3 measured on the training warps; OFF by default for compatibility, turn it ON for new work."}],"depth":["IMAGE",{"tooltip":"Depth map for the same frames, from a Depth Anything V2 node. Brightness = depth; polarity is handled by invert_depth. Leave unconnected if you are feeding moge_geometry instead."}],"moge_geometry":["MOGE_GEOMETRY",{"tooltip":"Metric geometry from Run MoGe Inference, used instead of the depth image. Depth then arrives in real metres, so depth_ratio, invert_depth and smooth_depth no longer apply, and pivot and distance become real distances. Measured against ground truth it roughly halves the warp error. hfov comes from its intrinsics when hfov is 0."}],"camera_info":["LOAD3D_CAMERA",{"tooltip":"Exact target camera, from Create Camera Info or a Load 3D viewport. When connected the pose comes from it, and the orbit controls - angles, distance, hfov, the pivot and the keyframe path - are all ignored."}],"preview_size":["INT",{"default":512,"min":128,"max":768,"step":32,"tooltip":"Longest side the clip is cached at for the live preview. The warp is scale-invariant, so this trades sharpness and memory for frame rate: about 1 MB per frame and 18 fps at 384, 1.8 MB and 11 fps at 512. Takes effect on the next run."}],"frame_index":["INT",{"default":1,"min":1,"max":100000,"step":1,"tooltip":"Playhead for the live preview, counted from 1. The scrub bar under the preview drives it; it does not affect the output and needs no re-run."}]},"hidden":{"unique_id":"UNIQUE_ID"}},"input_order":{"required":["frames","azimuth","elevation","distance","hfov","vertical_shift","depth_ratio","smooth_depth","invert_depth"],"optional":["roll_lock","pivot_override","pivot_x","pivot_y","pivot_z","use_keyframes","frame_count","keyframes","interp_motion","interpolation","keep_source_aim","depth","moge_geometry","camera_info","preview_size","frame_index"],"hidden":["unique_id"]},"is_input_list":false,"output":["IMAGE","IMAGE"],"output_is_list":[false,false],"output_name":["warp","orbit_view"],"name":"CrossViewWarp","display_name":"CrossView Warp","description":"","python_module":"custom_nodes.ComfyUI-CrossViewWarp","category":"CrossView","output_node":true,"has_intermediate_output":false,"output_tooltips":["The warp control video (magenta = disoccluded holes) -> connect to the IC-LoRA reference guide for the WARP stream.","Front view of the orbit globe with the camera marker at the requested az/el/dist - wire to a PreviewImage/SaveImage to document the camera setup."],"search_aliases":[]},"CustomCombo":{"input":{"required":{"choice":["COMBO",{"multiselect":false,"options":[]}]}},"input_order":{"required":["choice"]},"is_input_list":false,"output":["STRING","INT"],"output_is_list":[false,false],"output_name":["STRING","INDEX"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"CustomCombo","display_name":"Custom Combo","description":"","python_module":"comfy_extras.nodes_logic","category":"utilities","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptyAudio":{"input":{"required":{"duration":["FLOAT",{"tooltip":"Duration of the empty audio clip in seconds","default":60.0,"min":0.0,"max":18446744073709551615,"step":0.01}],"sample_rate":["INT",{"tooltip":"Sample rate of the empty audio clip.","advanced":true,"default":44100,"min":1,"max":192000}],"channels":["INT",{"tooltip":"Number of audio channels (1 for mono, 2 for stereo).","advanced":true,"default":2,"min":1,"max":2}]}},"input_order":{"required":["duration","sample_rate","channels"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyAudio","display_name":"Empty Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["blank audio"],"essentials_category":null,"has_intermediate_output":false},"EmptyFlux2LatentImage":{"input":{"required":{"width":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["width","height","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyFlux2LatentImage","display_name":"Empty Flux 2 Latent","description":"","python_module":"comfy_extras.nodes_flux","category":"model/latent/flux","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptyImage":{"input":{"required":{"width":["INT",{"default":512,"min":1,"max":16384,"step":1}],"height":["INT",{"default":512,"min":1,"max":16384,"step":1}],"batch_size":["INT",{"default":1,"min":1,"max":4096}],"color":["INT",{"default":0,"min":0,"max":16777215,"step":1,"display":"color"}]}},"input_order":{"required":["width","height","batch_size","color"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"EmptyImage","display_name":"Empty Image","description":"","python_module":"nodes","category":"image","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"EmptyLTXVLatentVideo":{"input":{"required":{"width":["INT",{"default":768,"min":64,"max":16384,"step":32}],"height":["INT",{"default":512,"min":64,"max":16384,"step":32}],"length":["INT",{"default":97,"min":1,"max":16384,"step":8}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["width","height","length","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyLTXVLatentVideo","display_name":"Empty LTXV Latent Video","description":"","python_module":"comfy_extras.nodes_lt","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptyLatentAudio":{"input":{"required":{"seconds":["FLOAT",{"default":47.6,"min":1.0,"max":1000.0,"step":0.1}],"batch_size":["INT",{"tooltip":"The number of latent images in the batch.","default":1,"min":1,"max":4096}]}},"input_order":{"required":["seconds","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyLatentAudio","display_name":"Empty Latent Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":"Audio","has_intermediate_output":false},"EmptyLatentImage":{"input":{"required":{"width":["INT",{"default":1024,"min":16,"max":16384,"step":8,"tooltip":"The width of the latent images in pixels."}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":8,"tooltip":"The height of the latent images in pixels."}],"batch_size":["INT",{"default":1,"min":1,"max":4096,"tooltip":"The number of latent images in the batch."}]}},"input_order":{"required":["width","height","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"EmptyLatentImage","display_name":"Empty Latent Image","description":"Create a new batch of empty latent images to be denoised via sampling.","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"output_tooltips":["The empty latent image batch."],"search_aliases":["empty","empty latent","new latent","create latent","blank latent","blank"]},"EmptyMiniMaxMusic3LatentAudio":{"input":{"required":{"seconds":["FLOAT",{"default":120.0,"min":0.04,"max":360.0,"step":0.04}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["seconds","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyMiniMaxMusic3LatentAudio","display_name":"Empty MiniMax Music3 Latent Audio","description":"Creates an empty MiniMax Music3 audio latent for the requested duration.","python_module":"comfy_extras.nodes_minimax_music","category":"model/latent/minimax music","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptySD3LatentImage":{"input":{"required":{"width":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["width","height","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptySD3LatentImage","display_name":null,"description":"","python_module":"comfy_extras.nodes_sd3","category":"model/latent/stable diffusion","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"Flux3LayoutPrompt":{"input":{"required":{"prompt":["STRING",{"tooltip":"The prompt text. Cite box N as <obj_N> to tie words to it.","multiline":true,"dynamicPrompts":true}],"element":["ARRAY",{"tooltip":"Prompt elements from the node Create Bounding Boxes. Colour palettes are not used, describe colours in the description."}]}},"input_order":{"required":["prompt","element"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["prompt"],"output_tooltips":[null],"output_matchtypes":null,"name":"Flux3LayoutPrompt","display_name":"FLUX 3 Image Layout Prompt","description":"Appends the boxes from Create Bounding Boxes to a FLUX 3 Image prompt as its layout table.","python_module":"comfy_extras.nodes_flux","category":"model/conditioning/flux","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"FluxKontextImageScale":{"input":{"required":{"image":["IMAGE",{}]}},"input_order":{"required":["image"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"FluxKontextImageScale","display_name":null,"description":"This node resizes the image to one that is more optimal for flux kontext.","python_module":"comfy_extras.nodes_flux","category":"model/conditioning/flux","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"FluxKontextMultiReferenceLatentMethod":{"input":{"required":{"conditioning":["CONDITIONING",{}],"reference_latents_method":["COMBO",{"advanced":true,"multiselect":false,"options":["offset","index","uxo/uno","index_timestep_zero"]}]}},"input_order":{"required":["conditioning","reference_latents_method"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"output_tooltips":[null],"output_matchtypes":null,"name":"FluxKontextMultiReferenceLatentMethod","display_name":"Edit Model Reference Method","description":"","python_module":"comfy_extras.nodes_flux","category":"model/conditioning/flux","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"FrameInterpolate":{"input":{"required":{"interp_model":["INTERP_MODEL",{}],"images":["IMAGE",{}],"multiplier":["INT",{"default":2,"min":2,"max":16}]}},"input_order":{"required":["interp_model","images","multiplier"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"FrameInterpolate","display_name":"Run Frame Interpolation Model","description":"","python_module":"comfy_extras.nodes_frame_interpolation","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["rife","film","frame interpolation","slow motion","interpolate frames","vfi"],"essentials_category":null,"has_intermediate_output":false},"FrameInterpolationModelLoader":{"input":{"required":{"model_name":["COMBO",{"tooltip":"Select a frame interpolation model to load. Models must be placed in the 'frame_interpolation' folder.","multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["model_name"]},"is_input_list":false,"output":["INTERP_MODEL"],"output_is_list":[false],"output_name":["INTERP_MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"FrameInterpolationModelLoader","display_name":"Load Frame Interpolation Model","description":"","python_module":"comfy_extras.nodes_frame_interpolation","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"GemmaAPITextEncode":{"input":{"required":{"api_key":["STRING",{"default":"","placeholder":"API_KEY","multiline":false,"tooltip":"API key for authentication"}],"prompt":["STRING",{"multiline":true,"default":"","tooltip":"Text prompt to encode"}],"enhance_prompt":["BOOLEAN",{"default":true,"tooltip":"When enabled, the prompt is enhanced using Gemma 3 before encoding"}],"ckpt_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"],{"tooltip":"The model to read the API text-encoder selector from. Either a checkpoint or a diffusion model."}]}},"input_order":{"required":["api_key","prompt","enhance_prompt","ckpt_name"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["conditioning"],"name":"GemmaAPITextEncode","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Gemma API Text Encode","description":"","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"api node/text/Lightricks","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"GetImageSize":{"input":{"required":{"image":["IMAGE",{}]},"hidden":{"unique_id":["UNIQUE_ID"]}},"input_order":{"required":["image"],"hidden":["unique_id"]},"is_input_list":false,"output":["INT","INT","INT"],"output_is_list":[false,false,false],"output_name":["width","height","batch_size"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"GetImageSize","display_name":"Get Image Size","description":"Returns the width, height, and batch size of the image.","python_module":"comfy_extras.nodes_images","category":"image","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["dimensions","resolution","image info"],"essentials_category":null,"has_intermediate_output":false},"GetVideoComponents":{"input":{"required":{"video":["VIDEO",{"tooltip":"The video to extract components from."}]}},"input_order":{"required":["video"]},"is_input_list":false,"output":["IMAGE","AUDIO","FLOAT","COMBO","COMBO"],"output_is_list":[false,false,false,false,false],"output_name":["images","audio","fps","bit_depth","color_space"],"output_tooltips":[null,null,null,null,null],"output_matchtypes":null,"name":"GetVideoComponents","display_name":"Get Video Components","description":"Extracts video frames, audio, frame rate, bit depth, and color space.","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["extract frames","split video","video to images","demux"],"essentials_category":null,"has_intermediate_output":false},"ImageColorSpace":{"input":{"required":{"image":["IMAGE",{}],"source":["COMBO",{"tooltip":"Color space of the input pixels.","default":"sRGB","multiselect":false,"options":["sRGB","HDR","HDR PQ","linear","HDR LogC3","HDR ACEScct"]}],"destination":["COMBO",{"tooltip":"Color space of the output pixels. Set the save node to this same color space. Convert LogC3 or ACEScct to linear before saving EXR, or to sRGB/HDR/HDR PQ before saving video.","default":"sRGB","multiselect":false,"options":["sRGB","HDR","HDR PQ","linear","HDR LogC3","HDR ACEScct"]}]}},"input_order":{"required":["image","source","destination"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageColorSpace","display_name":"Convert Image Color Space","description":"Convert sRGB, linear Rec.709, HDR (Rec.2020 HLG), HDR PQ (Rec.2020 PQ), HDR LogC3, and HDR ACEScct. LogC3 uses the EI 800 curve with Rec.709 primaries and codes clamped to [0, 1]. ACEScct uses AP1 primaries and D60 white, with Bradford adaptation to D65. Convert LogC3 or ACEScct to linear for EXR saving. Linear 1.0 uses the same 203-nit reference white as sRGB; HLG uses a 1000-nit reference display. Linear and ACEScct outputs preserve extended values. Linear-to-HDR conversions preserve highlights without tone mapping; HLG and PQ outputs clip negative channels. SDR output and PQ-to-HLG conversion tone-map excess luminance across the batch and compress out-of-gamut colors. Conversions compute in float32 and return the intermediate device and dtype. Straight alpha is not color-transformed.","python_module":"comfy_extras.nodes_images","category":"image/color","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"ImageCompare":{"input":{"required":{"compare_view":["IMAGECOMPARE",{"socketless":true}]},"optional":{"image_a":["IMAGE",{}],"image_b":["IMAGE",{}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["compare_view"],"optional":["image_a","image_b"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":[],"output_is_list":[],"output_name":[],"output_tooltips":[],"output_matchtypes":null,"name":"ImageCompare","display_name":"Compare Images","description":"Compares two images side by side with a slider.","python_module":"comfy_extras.nodes_image_compare","category":"image","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":"Image Tools","has_intermediate_output":false},"ImageFromBatch":{"input":{"required":{"image":["IMAGE",{}],"batch_index":["INT",{"default":0,"min":-16384,"max":16384}],"length":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["image","batch_index","length"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageFromBatch","display_name":"Get Image from Batch","description":"","python_module":"comfy_extras.nodes_images","category":"image/batch","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["select image","pick from batch","extract image"],"essentials_category":null,"has_intermediate_output":false},"ImageInvert":{"input":{"required":{"image":["IMAGE"]}},"input_order":{"required":["image"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"ImageInvert","display_name":"Invert Image Colors","description":"","python_module":"nodes","category":"image/color","output_node":false,"has_intermediate_output":false,"search_aliases":["reverse colors"],"essentials_category":"Image Tools"},"ImagePadForOutpaint":{"input":{"required":{"image":["IMAGE"],"left":["INT",{"default":0,"min":0,"max":16384,"step":8}],"top":["INT",{"default":0,"min":0,"max":16384,"step":8}],"right":["INT",{"default":0,"min":0,"max":16384,"step":8}],"bottom":["INT",{"default":0,"min":0,"max":16384,"step":8}],"feathering":["INT",{"default":40,"min":0,"max":16384,"step":1,"advanced":true}]}},"input_order":{"required":["image","left","top","right","bottom","feathering"]},"is_input_list":false,"output":["IMAGE","MASK"],"output_is_list":[false,false],"output_name":["IMAGE","MASK"],"name":"ImagePadForOutpaint","display_name":"Pad Image for Outpainting","description":"","python_module":"nodes","category":"image/transform","output_node":false,"has_intermediate_output":false,"search_aliases":["extend canvas","expand image"]},"ImageScale":{"input":{"required":{"image":["IMAGE"],"upscale_method":[["nearest-exact","bilinear","area","bicubic","lanczos"]],"width":["INT",{"default":512,"min":0,"max":16384,"step":1}],"height":["INT",{"default":512,"min":0,"max":16384,"step":1}],"crop":[["disabled","center"]]}},"input_order":{"required":["image","upscale_method","width","height","crop"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"ImageScale","display_name":"Upscale Image","description":"","python_module":"nodes","category":"image/upscaling","output_node":false,"has_intermediate_output":false,"search_aliases":["resize","resize image","scale image","image resize","zoom","zoom in","change size"],"essentials_category":"Image Tools"},"ImageStitch":{"input":{"required":{"image1":["IMAGE",{}],"direction":["COMBO",{"default":"right","multiselect":false,"options":["right","down","left","up"]}],"match_image_size":["BOOLEAN",{"default":true}],"spacing_width":["INT",{"advanced":true,"default":0,"min":0,"max":1024,"step":2}],"spacing_color":["COMBO",{"advanced":true,"default":"white","multiselect":false,"options":["white","black","red","green","blue"]}]},"optional":{"image2":["IMAGE",{}]}},"input_order":{"required":["image1","direction","match_image_size","spacing_width","spacing_color"],"optional":["image2"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageStitch","display_name":"Stitch Images","description":"Stitches image2 to image1 in the specified direction.\nIf image2 is not provided, returns image1 unchanged.\nOptional spacing can be added between images.","python_module":"comfy_extras.nodes_images","category":"image/transform","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["combine images","join images","concatenate images","side by side"],"essentials_category":null,"has_intermediate_output":false},"ImageToMask":{"input":{"required":{"image":["IMAGE",{}],"channel":["COMBO",{"multiselect":false,"options":["red","green","blue","alpha"]}]}},"input_order":{"required":["image","channel"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["MASK"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageToMask","display_name":"Convert Image to Mask","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["extract channel","channel to mask"],"essentials_category":null,"has_intermediate_output":false},"InvertMask":{"input":{"required":{"mask":["MASK",{}]}},"input_order":{"required":["mask"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["MASK"],"output_tooltips":[null],"output_matchtypes":null,"name":"InvertMask","display_name":"Invert Mask","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["reverse mask","flip mask"],"essentials_category":null,"has_intermediate_output":false},"JoinImageWithAlpha":{"input":{"required":{"image":["IMAGE",{}],"alpha":["MASK",{}]}},"input_order":{"required":["image","alpha"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"JoinImageWithAlpha","display_name":"Join Image with Alpha","description":"","python_module":"comfy_extras.nodes_compositing","category":"image/compositing","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["add transparency","apply alpha","composite alpha","RGBA"],"essentials_category":null,"has_intermediate_output":false},"JsonExtractString":{"input":{"required":{"json_string":["STRING",{"multiline":true}],"key":["STRING",{"multiline":false}]}},"input_order":{"required":["json_string","key"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"JsonExtractString","display_name":"Extract Text from JSON","description":"Extract a key's value from the first valid JSON object in the text, including Markdown code blocks.","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["json","extract json","parse json","json value","read json"],"essentials_category":null,"has_intermediate_output":false},"KSampler":{"input":{"required":{"model":["MODEL",{"tooltip":"The model used for denoising the input latent."}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true,"tooltip":"The random seed used for creating the noise."}],"steps":["INT",{"default":20,"min":1,"max":10000,"tooltip":"The number of steps used in the denoising process."}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01,"tooltip":"The Classifier-Free Guidance scale balances creativity and adherence to the prompt. Higher values result in images more closely matching the prompt however too high values will negatively impact quality."}],"sampler_name":[["euler","euler_cfg_pp","euler_ancestral","euler_ancestral_cfg_pp","heun","heunpp2","exp_heun_2_x0","exp_heun_2_x0_sde","dpm_2","dpm_2_ancestral","lms","dpm_fast","dpm_adaptive","dpmpp_2s_ancestral","dpmpp_2s_ancestral_cfg_pp","dpmpp_sde","dpmpp_sde_gpu","dpmpp_2m","dpmpp_2m_cfg_pp","dpmpp_2m_sde","dpmpp_2m_sde_gpu","dpmpp_2m_sde_heun","dpmpp_2m_sde_heun_gpu","dpmpp_3m_sde","dpmpp_3m_sde_gpu","ddpm","lcm","ipndm","ipndm_v","deis","cfgpp_ud10_ab","res_multistep","res_multistep_cfg_pp","res_multistep_ancestral","res_multistep_ancestral_cfg_pp","gradient_estimation","gradient_estimation_cfg_pp","er_sde","seeds_2","seeds_3","sa_solver","sa_solver_pece","ddim","uni_pc","uni_pc_bh2"],{"tooltip":"The algorithm used when sampling, this can affect the quality, speed, and style of the generated output."}],"scheduler":[["simple","sgm_uniform","karras","exponential","ddim_uniform","beta","normal","linear_quadratic","kl_optimal"],{"tooltip":"The scheduler controls how noise is gradually removed to form the image."}],"positive":["CONDITIONING",{"tooltip":"The conditioning describing the attributes you want to include in the image."}],"negative":["CONDITIONING",{"tooltip":"The conditioning describing the attributes you want to exclude from the image."}],"latent_image":["LATENT",{"tooltip":"The latent image to denoise."}],"denoise":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01,"tooltip":"The amount of denoising applied, lower values will maintain the structure of the initial image allowing for image to image sampling."}]}},"input_order":{"required":["model","seed","steps","cfg","sampler_name","scheduler","positive","negative","latent_image","denoise"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"KSampler","display_name":"KSampler","description":"Uses the provided model, positive and negative conditioning to denoise the latent image.","python_module":"nodes","category":"model/sampling","output_node":false,"has_intermediate_output":false,"output_tooltips":["The denoised latent."],"search_aliases":["sampler","sample","generate","denoise","diffuse","txt2img","img2img"]},"KSamplerAdvanced":{"input":{"required":{"model":["MODEL"],"add_noise":[["enable","disable"],{"advanced":true}],"noise_seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}],"steps":["INT",{"default":20,"min":1,"max":10000}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"sampler_name":[["euler","euler_cfg_pp","euler_ancestral","euler_ancestral_cfg_pp","heun","heunpp2","exp_heun_2_x0","exp_heun_2_x0_sde","dpm_2","dpm_2_ancestral","lms","dpm_fast","dpm_adaptive","dpmpp_2s_ancestral","dpmpp_2s_ancestral_cfg_pp","dpmpp_sde","dpmpp_sde_gpu","dpmpp_2m","dpmpp_2m_cfg_pp","dpmpp_2m_sde","dpmpp_2m_sde_gpu","dpmpp_2m_sde_heun","dpmpp_2m_sde_heun_gpu","dpmpp_3m_sde","dpmpp_3m_sde_gpu","ddpm","lcm","ipndm","ipndm_v","deis","cfgpp_ud10_ab","res_multistep","res_multistep_cfg_pp","res_multistep_ancestral","res_multistep_ancestral_cfg_pp","gradient_estimation","gradient_estimation_cfg_pp","er_sde","seeds_2","seeds_3","sa_solver","sa_solver_pece","ddim","uni_pc","uni_pc_bh2"]],"scheduler":[["simple","sgm_uniform","karras","exponential","ddim_uniform","beta","normal","linear_quadratic","kl_optimal"]],"positive":["CONDITIONING"],"negative":["CONDITIONING"],"latent_image":["LATENT"],"start_at_step":["INT",{"default":0,"min":0,"max":10000,"advanced":true}],"end_at_step":["INT",{"default":10000,"min":0,"max":10000,"advanced":true}],"return_with_leftover_noise":[["disable","enable"],{"advanced":true}]}},"input_order":{"required":["model","add_noise","noise_seed","steps","cfg","sampler_name","scheduler","positive","negative","latent_image","start_at_step","end_at_step","return_with_leftover_noise"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"KSamplerAdvanced","display_name":"KSampler (Advanced)","description":"","python_module":"nodes","category":"model/sampling","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"KSamplerSelect":{"input":{"required":{"sampler_name":["COMBO",{"multiselect":false,"options":["euler","euler_cfg_pp","euler_ancestral","euler_ancestral_cfg_pp","heun","heunpp2","exp_heun_2_x0","exp_heun_2_x0_sde","dpm_2","dpm_2_ancestral","lms","dpm_fast","dpm_adaptive","dpmpp_2s_ancestral","dpmpp_2s_ancestral_cfg_pp","dpmpp_sde","dpmpp_sde_gpu","dpmpp_2m","dpmpp_2m_cfg_pp","dpmpp_2m_sde","dpmpp_2m_sde_gpu","dpmpp_2m_sde_heun","dpmpp_2m_sde_heun_gpu","dpmpp_3m_sde","dpmpp_3m_sde_gpu","ddpm","lcm","ipndm","ipndm_v","deis","cfgpp_ud10_ab","res_multistep","res_multistep_cfg_pp","res_multistep_ancestral","res_multistep_ancestral_cfg_pp","gradient_estimation","gradient_estimation_cfg_pp","er_sde","seeds_2","seeds_3","sa_solver","sa_solver_pece","ddim","uni_pc","uni_pc_bh2"]}]}},"input_order":{"required":["sampler_name"]},"is_input_list":false,"output":["SAMPLER"],"output_is_list":[false],"output_name":["SAMPLER"],"output_tooltips":[null],"output_matchtypes":null,"name":"KSamplerSelect","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/samplers","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXAddVideoICLoRAGuide":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"latent":["LATENT",{"tooltip":"Video-only latent to condition. Must be a 5D video latent (batch, channels, frames, height, width)."}],"image":["IMAGE",{}],"frame_idx":["INT",{"tooltip":"RoPE position of the reference, in pixel frames.\n0 aligns the reference with the generated clip. Use for video references.\n>0 places the reference at a specific pixel-frame position in the clip.\n-1 places the reference outside the generated frame range, while still using it for attention conditioning. Use for single-image references.\nFor multi-frame references, >0 values are rounded down to 1 mod 8.","default":0,"min":-9999,"max":9999}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}],"latent_downscale_factor":["FLOAT",{"tooltip":"For IC-LoRA on small grid. 1 means original size, 2 means half size, 3 means third, etc.","default":1.0,"min":1.0,"max":10.0,"step":1.0}],"crop":["COMBO",{"tooltip":"Crop mode when resizing. 'center' crops to fit, 'disabled' stretches to fit.","default":"disabled","multiselect":false,"options":["disabled","center"]}],"use_tiled_encode":["BOOLEAN",{"tooltip":"Enable tiled VAE encoding for large resolutions/long videos to reduce memory usage.","default":false}],"tile_size":["INT",{"tooltip":"Spatial tile size for tiled encoding. Only used when use_tiled_encode is enabled.","default":256,"min":64,"max":512,"step":32}],"tile_overlap":["INT",{"tooltip":"Overlap between tiles for tiled encoding. Only used when use_tiled_encode is enabled.","default":64,"min":16,"max":256,"step":16}]},"optional":{"use_streaming":["BOOLEAN",{"tooltip":"Off (default): encode the whole clip as one IC-LoRA guide. On: re-encode one fresh guide per LTXVTiledFusionSampler temporal window and write the plan the sampler reads. Set tile_frames on this node and the sampler to the same value (97 for LTX).","default":false}],"tile_frames":["INT",{"tooltip":"Window length in pixel frames, used only when use_streaming is on. Snapped to 8n+1. Must match LTXVTiledFusionSampler. Ignored when use_streaming is off.","default":97,"min":0,"max":1000,"step":1}]}},"input_order":{"required":["positive","negative","vae","latent","image","frame_idx","strength","latent_downscale_factor","crop","use_tiled_encode","tile_size","tile_overlap"],"optional":["use_streaming","tile_frames"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXAddVideoICLoRAGuide","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Add Video IC-LoRA Guide","description":"Adds one or more conditioning frames starting at the specified frame index. Supports both single images and multi-frame videos. The latent_downscale_factor resizes input to a fraction of the target size (1 = original, 2 = half, 3 = third, etc.) for IC-LoRA on small grids. For LTXVTiledFusionSampler temporal windows, turn on use_streaming and set tile_frames to the same 8n+1 value as the sampler (97 for LTX).","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXAddVideoICLoRAGuideAdvanced":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"latent":["LATENT",{"tooltip":"Video-only latent to condition. Must be a 5D video latent."}],"image":["IMAGE",{}],"frame_idx":["INT",{"tooltip":"RoPE position of the reference, in pixel frames.\n0 aligns the reference with the generated clip. Use for video references.\n>0 places the reference at a specific pixel-frame position in the clip.\n-1 places the reference outside the generated frame range, while still using it for attention conditioning. Use for single-image references.\nFor multi-frame references, >0 values are rounded down to 1 mod 8.","default":0,"min":-9999,"max":9999}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}],"latent_downscale_factor":["FLOAT",{"tooltip":"For IC-LoRA on small grid. 1 = original size, 2 = half, etc.","default":1.0,"min":1.0,"max":10.0,"step":1.0}],"crop":["COMBO",{"default":"disabled","multiselect":false,"options":["disabled","center"]}],"use_tiled_encode":["BOOLEAN",{"default":false}],"tile_size":["INT",{"default":256,"min":64,"max":512,"step":32}],"tile_overlap":["INT",{"default":64,"min":16,"max":256,"step":16}],"attention_strength":["FLOAT",{"tooltip":"Controls how strongly this guide influences generation via self-attention. 1.0 = full conditioning (default), 0.0 = ignore. When an attention_mask is also provided, this multiplies the mask values.","default":1.0,"min":0.0,"max":1.0,"step":0.01}]},"optional":{"attention_mask":["MASK",{"tooltip":"Optional pixel-space spatial mask. Shape (F, H, W) or (H, W). Values in [0, 1]. Controls per-region conditioning influence. Multiplied by attention_strength."}],"use_streaming":["BOOLEAN",{"tooltip":"Off (default): whole-clip guide. On: one fresh encode per fusion temporal window. Must match the sampler's tile_frames.","default":false}],"tile_frames":["INT",{"tooltip":"Used only when use_streaming is on. Snapped to 8n+1. 97 for LTX. Must match the sampler.","default":97,"min":0,"max":1000,"step":1}]}},"input_order":{"required":["positive","negative","vae","latent","image","frame_idx","strength","latent_downscale_factor","crop","use_tiled_encode","tile_size","tile_overlap","attention_strength"],"optional":["attention_mask","use_streaming","tile_frames"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXAddVideoICLoRAGuideAdvanced","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Add Video IC-LoRA Guide Advanced","description":"Adds IC-LoRA guide conditioning with per-guide attention strength control. Same as LTXAddVideoICLoRAGuide, but allows controlling how strongly this guide influences generation via self-attention, optionally with a spatial mask. use_streaming + matching tile_frames re-encodes one guide per Tiled Fusion window.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXFloatToInt":{"input":{"required":{"a":["FLOAT",{"default":0.0}]}},"input_order":{"required":["a"]},"is_input_list":false,"output":["INT"],"output_is_list":[false],"output_name":["INT"],"name":"LTXFloatToInt","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Float To Int","description":"","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"math/conversion","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LTXICLoRALoaderModelOnly":{"input":{"required":{"model":["MODEL",{}],"lora_name":["COMBO",{"multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}],"strength_model":["FLOAT",{"default":1.0,"min":-100.0,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","lora_name","strength_model"]},"is_input_list":false,"output":["MODEL","FLOAT"],"output_is_list":[false,false],"output_name":["model","latent_downscale_factor"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"LTXICLoRALoaderModelOnly","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 IC-LoRA Loader Model Only","description":"Loads a LoRA model and extracts the latent_downscale_factor from the safetensors metadata.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVAddGuide":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"latent":["LATENT",{}],"image":["IMAGE",{"tooltip":"Image or video to condition the latent video on. Must be 8*n + 1 frames. If the video is not 8*n + 1 frames, it will be cropped to the nearest 8*n + 1 frames."}],"frame_idx":["INT",{"tooltip":"Frame index to start the conditioning at. For single-frame images or videos with 1-8 frames, any frame_idx value is acceptable. For videos with 9+ frames, frame_idx must be divisible by 8, otherwise it will be rounded down to the nearest multiple of 8. Negative values are counted from the end of the video.","default":0,"min":-9999,"max":9999}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":10.0,"step":0.01}]},"optional":{"attention_mask":["MASK",{"tooltip":"Optional pixel-space spatial mask. Controls per-region conditioning influence via self-attention, multiplied by strength."}],"iclora_parameters":["IC_LORA_PARAMETERS",{"tooltip":"Optional IC-LoRA parameters from a Get IC-LoRA Parameters node. Used for adjusting guide processing as required by certain IC-LoRAs (eg. those with a reference_downscale_factor > 1). When chained, each LTXVAddGuide uses only the parameters connected to it."}]}},"input_order":{"required":["positive","negative","vae","latent","image","frame_idx","strength"],"optional":["attention_mask","iclora_parameters"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXVAddGuide","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVAudioVAEDecode":{"input":{"required":{"samples":["LATENT",{"tooltip":"The latent to be decoded."}],"audio_vae":["VAE",{"display_name":"Audio VAE","tooltip":"The Audio VAE model used for decoding the latent."}]}},"input_order":{"required":["samples","audio_vae"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["Audio"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVAudioVAEDecode","display_name":"LTXV Audio VAE Decode","description":"","python_module":"comfy_extras.nodes_lt_audio","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVAudioVAEEncode":{"input":{"required":{"audio":["AUDIO",{"tooltip":"The audio to be encoded."}],"audio_vae":["VAE",{"display_name":"Audio VAE","tooltip":"The Audio VAE model to use for encoding."}]}},"input_order":{"required":["audio","audio_vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["Audio Latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVAudioVAEEncode","display_name":"LTXV Audio VAE Encode","description":"","python_module":"comfy_extras.nodes_lt_audio","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVConcatAVLatent":{"input":{"required":{"video_latent":["LATENT",{}],"audio_latent":["LATENT",{}]}},"input_order":{"required":["video_latent","audio_latent"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVConcatAVLatent","display_name":"Concat AV Latent","description":"Merge a video latent and an audio latent into a joint AV latent (any AV model, e.g. LTXV or MiniMax H3).","python_module":"comfy_extras.nodes_lt","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVConditioning":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"frame_rate":["FLOAT",{"default":25.0,"min":0.0,"max":1000.0,"step":0.01}]}},"input_order":{"required":["positive","negative","frame_rate"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING"],"output_is_list":[false,false],"output_name":["positive","negative"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"LTXVConditioning","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVCropGuides":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"latent":["LATENT",{}]}},"input_order":{"required":["positive","negative","latent"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXVCropGuides","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVDualCFGGuider":{"input":{"required":{"model":["MODEL",{}],"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"video_cfg":["FLOAT",{"default":3.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"audio_cfg":["FLOAT",{"default":7.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}]}},"input_order":{"required":["model","positive","negative","video_cfg","audio_cfg"]},"is_input_list":false,"output":["GUIDER"],"output_is_list":[false],"output_name":["GUIDER"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVDualCFGGuider","display_name":"LTXV Dual CFG Guider","description":"Separate CFG scales for the video and audio modalities of a packed LTXV-AV latent.","python_module":"comfy_extras.nodes_lt","category":"model/sampling/guiders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVEmptyLatentAudio":{"input":{"required":{"frames_number":["INT",{"tooltip":"Number of frames.","default":97,"min":1,"max":1000,"step":1,"display":"number"}],"frame_rate":["FLOAT,INT",{"tooltip":"Number of frames per second.","default":25.0,"widgetType":"FLOAT","min":1.0,"max":1000.0,"step":0.01,"display":"number"}],"batch_size":["INT",{"tooltip":"The number of latent audio samples in the batch.","default":1,"min":1,"max":4096,"display":"number"}],"audio_vae":["VAE",{"display_name":"Audio VAE","tooltip":"The Audio VAE model to get configuration from."}]}},"input_order":{"required":["frames_number","frame_rate","batch_size","audio_vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["Latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVEmptyLatentAudio","display_name":"Empty LTXV Latent Audio","description":"","python_module":"comfy_extras.nodes_lt_audio","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVHDRDecodePostprocess":{"input":{"required":{"image":["IMAGE"]},"optional":{"transfer":[["acescct","logc3"],{"default":"acescct","tooltip":"Working-space curve of the VAE decode. Use acescct for SDR\u2192HDR IC-LoRA (after LTXVSDRToHDRWorkingSpace) and for native HDR EXR; logc3 only for legacy LogC3 HDR LoRA checkpoints."}],"exposure":["FLOAT",{"default":0.0,"min":-10.0,"max":10.0,"step":0.1,"display":"slider","tooltip":"Exposure in stops (EV). 0 = no change, +1 = 2x brighter, -1 = half brightness. Affects the tonemapped preview only."}],"save_exr":["BOOLEAN",{"default":false,"tooltip":"Save HDR frames as an EXR sequence."}],"exr_color_space":[["linear","acescct","srgb_linear","acescg"],{"default":"acescg","tooltip":"EXR contents when save_exr is on. Default 'acescg' is tagged ACEScg scene-linear. 'acescct' is ACEScct log; 'srgb_linear' is tagged Rec.709 linear; 'linear' is legacy raw decompress (untagged). All paths use OpenImageIO."}],"output_dir":["STRING",{"default":"output/hdr_exr","tooltip":"Directory for EXR frames (relative to ComfyUI output directory, or absolute path)."}],"filename_prefix":["STRING",{"default":"frame"}],"half_precision":["BOOLEAN",{"default":true,"tooltip":"Save EXR as float16 (half). Smaller files, negligible quality loss for most workflows."}]}},"input_order":{"required":["image"],"optional":["transfer","exposure","save_exr","exr_color_space","output_dir","filename_prefix","half_precision"]},"is_input_list":false,"output":["IMAGE","IMAGE"],"output_is_list":[false,false],"output_name":["tonemapped","hdr_linear"],"name":"LTXVHDRDecodePostprocess","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXVHDR Decode Postprocess","description":"Decompresses VAE-decoded HDR (ACEScct or legacy LogC3), tonemaps for preview, and optionally writes EXR (linear / ACEScct / sRGB linear / ACEScg). Wire hdr_linear into LTXVSaveHLG for a BT.2020/HLG master.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":true,"has_intermediate_output":false,"search_aliases":[]},"LTXVImgToVideoConditionOnly":{"input":{"required":{"vae":["VAE"],"image":["IMAGE"],"latent":["LATENT"],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0}]},"optional":{"bypass":["BOOLEAN",{"default":false,"tooltip":"Bypass the conditioning."}]}},"input_order":{"required":["vae","image","latent","strength"],"optional":["bypass"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"name":"LTXVImgToVideoConditionOnly","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Img To Video Condition Only","description":"Applies image conditioning to the first frames of an existing latent. Creates a noise mask to control conditioning strength.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"conditioning/video_models","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LTXVImgToVideoInplace":{"input":{"required":{"vae":["VAE",{}],"image":["IMAGE",{}],"latent":["LATENT",{}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0}],"bypass":["BOOLEAN",{"tooltip":"Bypass the conditioning.","default":false}]}},"input_order":{"required":["vae","image","latent","strength","bypass"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVImgToVideoInplace","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVInpaintPreprocess":{"input":{"required":{"images":["IMAGE",{"tooltip":"Video frames to composite onto the green background."}],"mask":["MASK",{"tooltip":"Mask indicating regions to replace with green. Single-frame masks are broadcast."}]}},"input_order":{"required":["images","mask"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["image"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVInpaintPreprocess","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Inpaint Preprocess","description":"Composites images with a green (#66FF00) background where mask is active, for inpainting conditioning. Always paints SDR #66FF00 (not ACEScct-converted) \u2014 that key matches IC-LoRA training even on native HDR / ACEScct plates.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/image_processing","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVLaplacianPyramidBlend":{"input":{"required":{"image_a":["IMAGE",{"tooltip":"First source image."}],"image_b":["IMAGE",{"tooltip":"Second source image."}],"mask":["MASK",{"tooltip":"Blend mask (white = image_a, black = image_b)."}],"trim_to_shortest":["BOOLEAN",{"tooltip":"Trim image_a, image_b, and mask to the shortest sequence length before blending.","default":true}],"mask_low_res_dilation":["INT",{"tooltip":"Downscale the mask to long side 64, dilate it spatially, then resize it back before blending.","default":5,"min":0,"max":15}]}},"input_order":{"required":["image_a","image_b","mask","trim_to_shortest","mask_low_res_dilation"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["image"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVLaplacianPyramidBlend","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTX Laplacian Pyramid Blend","description":"Blend two images seamlessly using Laplacian pyramid blending.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/utility","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVLatentUpsampler":{"input":{"required":{"samples":["LATENT",{}],"upscale_model":["LATENT_UPSCALE_MODEL",{}],"vae":["VAE",{}]}},"input_order":{"required":["samples","upscale_model","vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVLatentUpsampler","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt_upsampler","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVLoadConditioning":{"input":{"required":{"file_name":["COMBO",{"multiselect":false,"options":[""]}],"device":["COMBO",{"multiselect":false,"options":["cpu","gpu"]}]}},"input_order":{"required":["file_name","device"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVLoadConditioning","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Load Conditioning","description":"Load precomputed text conditioning from models/embeddings. Supports Comfy conditioning_data_* files and pipeline scene embeddings (video_context / audio_context), e.g. HDR IC-LoRA.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"lightricks/LTXV","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVPreprocess":{"input":{"required":{"image":["IMAGE",{}],"img_compression":["INT",{"tooltip":"Amount of compression to apply on image.","default":35,"min":0,"max":100}]}},"input_order":{"required":["image","img_compression"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["output_image"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVPreprocess","display_name":"LTXV Preprocess","description":"","python_module":"comfy_extras.nodes_lt","category":"video/preprocessors","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVSDRToHDRWorkingSpace":{"input":{"required":{"image":["IMAGE"],"color_space":[["srgb","srgb_gamma","acescg","acescct"],{"default":"srgb_gamma","tooltip":"SDR ingest IDT. srgb_gamma = display sRGB (EOTF then ACEScct); srgb = scene-linear Rec.709; acescg / acescct for float EXR-style plates."}]}},"input_order":{"required":["image","color_space"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["acescct"],"name":"LTXVSDRToHDRWorkingSpace","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXVSDR To HDR Working Space","description":"SDR \u2192 ACEScct working space for HDR IC-LoRA (pipelines SDRColorSpace). Wire LoadVideo images through this before resize / IC-LoRA guide.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LTXVSaveHLG":{"input":{"required":{"hdr_linear":["IMAGE",{"tooltip":"Scene-linear HDR from LTXVHDRDecodePostprocess. With transfer=acescct this is ACEScg linear \u2014 leave linear_primaries=acescg. For logc3, use rec709."}],"frame_rate":["FLOAT",{"default":24.0,"min":1.0,"max":120.0,"step":0.001}],"filename_prefix":["STRING",{"default":"hdr/ltxv_hlg"}]},"optional":{"linear_primaries":[["acescg","rec709"],{"default":"acescg","tooltip":"Primaries of hdr_linear. ACEScct decode (SDR\u2192HDR IC-LoRA or native HDR) \u2192 acescg. Legacy LogC3 \u2192 rec709."}],"audio":["AUDIO",{"tooltip":"Optional track to mux into the HLG MP4 (e.g. Decode AUDIO). Omit for video-only."}]}},"input_order":{"required":["hdr_linear","frame_rate","filename_prefix"],"optional":["linear_primaries","audio"]},"is_input_list":false,"output":[],"output_is_list":[],"output_name":[],"name":"LTXVSaveHLG","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Save HLG","description":"Writes a BT.2020/HLG 10-bit HEVC master from scene-linear hdr_linear (from LTXVHDRDecodePostprocess). Maps linear_primaries \u2192 Rec.2020 directly (no Rec.709 hop), then HLG. Optional AUDIO is muxed as AAC.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":true,"has_intermediate_output":false,"search_aliases":[]},"LTXVSeparateAVLatent":{"input":{"required":{"av_latent":["LATENT",{}]}},"input_order":{"required":["av_latent"]},"is_input_list":false,"output":["LATENT","LATENT"],"output_is_list":[false,false],"output_name":["video_latent","audio_latent"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"LTXVSeparateAVLatent","display_name":"Separate AV Latent","description":"Split a joint AV latent into its video and audio latents (any AV model, e.g. LTXV or MiniMax H3).","python_module":"comfy_extras.nodes_lt","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVSetAudioRefTokens":{"input":{"required":{"positive":["CONDITIONING",{"tooltip":"Positive conditioning to attach the reference audio tokens to."}],"negative":["CONDITIONING",{"tooltip":"Negative conditioning to attach the reference audio tokens to."}],"audio_latent":["LATENT",{"tooltip":"Encoded audio latent from LTXV Audio VAE Encode."}]}},"input_order":{"required":["positive","negative","audio_latent"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","frozen_audio"],"output_tooltips":["Positive conditioning with reference audio tokens attached.","Negative conditioning with reference audio tokens attached.","Audio latent with noise_mask=0, fully frozen during denoising."],"output_matchtypes":null,"name":"LTXVSetAudioRefTokens","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Set Audio Ref Tokens","description":"Provides speaker identity context for audio generation by attaching reference audio tokens to the conditioning. The tokens are prepended with negative temporal positions so the model treats them as context rather than generation targets.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVVAEForceFloat32":{"input":{"required":{"vae":["VAE"]}},"input_order":{"required":["vae"]},"is_input_list":false,"output":["VAE"],"output_is_list":[false],"output_name":["VAE"],"name":"LTXVVAEForceFloat32","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXVVAE Force Float32","description":"Forces float32 on the video VAE (vae_dtype + first_stage_model) for native HDR. IMPORTANT: this does NOT clone a float32 VAE for the output \u2014 it mutates the input object in place and returns the same instance. If you also wire that loader's VAE (or this output) into another branch, those nodes share the mutated model: they may see float16 or float32 depending on execution order (a race with the original dtype). Prefer a single chain VAELoader \u2192 this node \u2192 all encode/decode / IC-LoRA guide uses.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LatentUpscaleModelLoader":{"input":{"required":{"model_name":["COMBO",{"multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["model_name"]},"is_input_list":false,"output":["LATENT_UPSCALE_MODEL"],"output_is_list":[false],"output_name":["LATENT_UPSCALE_MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"LatentUpscaleModelLoader","display_name":"Load Latent Upscale Model","description":"","python_module":"comfy_extras.nodes_hunyuan","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LoadAudio":{"input":{"required":{"audio":["COMBO",{"multiselect":false,"options":["clip.mp4","drive_doll_street_dance.mp4","grainy_perfume_shot_crf32.mp4","ltx_23_audio.mp3","stone_ruins.mp4"],"audio_upload":true}]}},"input_order":{"required":["audio"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadAudio","display_name":"Load Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["import audio","open audio","audio file"],"essentials_category":"Audio","has_intermediate_output":false},"LoadBackgroundRemovalModel":{"input":{"required":{"bg_removal_name":["COMBO",{"tooltip":"The model used to remove backgrounds from images","multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["bg_removal_name"]},"is_input_list":false,"output":["BACKGROUND_REMOVAL"],"output_is_list":[false],"output_name":["bg_model"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadBackgroundRemovalModel","display_name":"Load Background Removal Model","description":"","python_module":"comfy_extras.nodes_bg_removal","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LoadImage":{"input":{"required":{"image":[["cactus_man.png","example.png","fashion_portrait_blue_silk_gown_vis.png","lab_smoke-invert-chain_smoke_00005_.png","ltx23_flf2v_first_frame.png","ltx23_flf2v_last_frame.png","neon_cyborg_portrait.png","neon_guitarist.png","portrait.png","ref_doll_realistic.png","robot_hand.png","robot_hand_energy.png","server_out.png","smoke_00005_.png","the_forgotten_gate.png","the_lily_veil.png","watch_macro_shot.png","woman_gold_earring_teal_top_portrait.png"],{"image_upload":true}]}},"input_order":{"required":["image"]},"is_input_list":false,"output":["IMAGE","MASK"],"output_is_list":[false,false],"output_name":["IMAGE","MASK"],"name":"LoadImage","display_name":"Load Image","description":"","python_module":"nodes","category":"image","output_node":false,"has_intermediate_output":false,"search_aliases":["load image","open image","import image","image input","upload image","read image","image loader"],"essentials_category":"Basics"},"LoadMoGeModel":{"input":{"required":{"model_name":["COMBO",{"multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["model_name"]},"is_input_list":false,"output":["MOGE_MODEL"],"output_is_list":[false],"output_name":["MOGE_MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadMoGeModel","display_name":"Load MoGe Model","description":"","python_module":"comfy_extras.nodes_moge","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LoadVideo":{"input":{"required":{"file":["COMBO",{"multiselect":false,"options":["clip.mp4","drive_doll_street_dance.mp4","grainy_perfume_shot_crf32.mp4","stone_ruins.mp4"],"video_upload":true}]}},"input_order":{"required":["file"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["VIDEO"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadVideo","display_name":"Load Video","description":"","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["import video","open video","video file"],"essentials_category":"Basics","has_intermediate_output":true},"LoraLoaderModelOnly":{"input":{"required":{"model":["MODEL"],"lora_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]],"strength_model":["FLOAT",{"default":1.0,"min":-100.0,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","lora_name","strength_model"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"LoraLoaderModelOnly","display_name":"Load LoRA","description":"This LoRAs loader is used to modify the diffusion model, altering the way in which latents are denoised such as applying styles. Multiple LoRA nodes can be linked together.","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"output_tooltips":["The modified diffusion model.","The modified CLIP model."],"search_aliases":["lora","load lora","apply lora","lora loader","lora model"],"essentials_category":"Image Generation"},"ManualSigmas":{"input":{"required":{"sigmas":["STRING",{"default":"1, 0.5","multiline":false}]}},"input_order":{"required":["sigmas"]},"is_input_list":false,"output":["SIGMAS"],"output_is_list":[false],"output_name":["SIGMAS"],"output_tooltips":[null],"output_matchtypes":null,"name":"ManualSigmas","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/sigmas","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["custom noise schedule","define sigmas"],"essentials_category":null,"has_intermediate_output":false},"MaskPreview":{"input":{"required":{"mask":["MASK",{}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["mask"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["mask"],"output_tooltips":[null],"output_matchtypes":null,"name":"MaskPreview","display_name":"Preview Mask","description":"Preview the masks without saving them to the ComfyUI output directory.","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["preview","preview mask","show mask","view mask","inspect mask","debug mask"],"essentials_category":null,"has_intermediate_output":false},"MaskToImage":{"input":{"required":{"mask":["MASK",{}]}},"input_order":{"required":["mask"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"MaskToImage","display_name":"Convert Mask to Image","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["convert mask"],"essentials_category":null,"has_intermediate_output":false},"MiniMaxH3AddGuide":{"input":{"required":{"positive":["CONDITIONING",{}],"latent":["LATENT",{}],"frame_idx":["INT",{"tooltip":"Frame index to anchor the image or the clip's first frame at. Negative values are counted from the end of the video.","default":0,"min":-9999,"max":9999}]},"optional":{"vae":["VAE",{"tooltip":"Video VAE, needed when an image is connected."}],"audio_vae":["VAE",{"tooltip":"Audio VAE, needed when an audio is connected."}],"image":["IMAGE",{"tooltip":"Image or video frames to anchor. Multi-frame batches are anchored as a clip and cropped down to the model's valid clip lengths: 5, 22, 39... (17k + 5) frames. Batches shorter than 5 frames use only the first image."}],"audio":["AUDIO",{"tooltip":"Soundtrack to anchor starting at the same frame index, cropped to the video's remaining duration."}]}},"input_order":{"required":["positive","latent","frame_idx"],"optional":["vae","audio_vae","image","audio"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["positive"],"output_tooltips":[null],"output_matchtypes":null,"name":"MiniMaxH3AddGuide","display_name":"Add Guide for MiniMax H3","description":"Anchor an image, a short clip, audio, or a clip with its soundtrack at any frame of a MiniMax H3 video. Chain several nodes to anchor several frames.","python_module":"comfy_extras.nodes_minimax_h3","category":"model/conditioning/minimax","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"MiniMaxH3ReferenceToVideo":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"multiline":true,"dynamicPrompts":true}],"width":["INT",{"default":1344,"min":32,"max":16384,"step":32}],"height":["INT",{"default":768,"min":32,"max":16384,"step":32}],"length":["INT",{"tooltip":"Frame count at 24 fps, (124 = ~5s, trained range is ~124-362)","default":124,"min":5,"max":3600,"step":17}],"ref_image_size":["COMBO",{"tooltip":"Reference image sizing. 'match' scales each ref (down only, keeping aspect) to the generation's pixel area; 'max' uses the reference pipeline's 2048px short edge for best identity fidelity. Reference tokens ride through every sampling step, so 'max' can be several times slower.","default":"match","multiselect":false,"options":["match","max"]}]},"optional":{"vae":["VAE",{"tooltip":"Video VAE. Without it reference images/videos only condition the text encoder."}],"audio_vae":["VAE",{"tooltip":"Audio VAE. Without it reference audio only conditions the text encoder."}],"ref_images":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_image":["IMAGE",{"tooltip":"Reference image (downscaled to 2048 short edge if larger, never upscaled)"}]}},"prefix":"ref_image_","min":0,"max":9}}],"ref_videos":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_video":["IMAGE",{"tooltip":"Reference video frames at 24 fps (2-15s)"}]}},"prefix":"ref_video_","min":0,"max":3}}],"ref_video_audios":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_video_audio":["AUDIO",{"tooltip":"Soundtrack of the same-numbered reference video"}]}},"prefix":"ref_video_audio_","min":0,"max":3}}],"ref_audios":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_audio":["AUDIO",{"tooltip":"Standalone reference audio"}]}},"prefix":"ref_audio_","min":0,"max":3}}]}},"input_order":{"required":["clip","prompt","width","height","length","ref_image_size"],"optional":["vae","audio_vae","ref_images","ref_videos","ref_video_audios","ref_audios"]},"is_input_list":false,"output":["CONDITIONING","LATENT"],"output_is_list":[false,false],"output_name":["positive","LATENT"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"MiniMaxH3ReferenceToVideo","display_name":"MiniMax H3 Reference to Video","description":"<Picture i> / <Video k> / <Audio j> reference conditioning for MiniMax H3. Use the same tags when prompting.","python_module":"comfy_extras.nodes_minimax_h3","category":"model/conditioning/minimax","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"MiniMaxH3SigmaShift":{"input":{"required":{"model":["MODEL",{}],"shift_video":["FLOAT",{"default":12.0,"min":0.01,"max":100.0,"step":0.01}],"shift_audio":["FLOAT",{"default":3.0,"min":0.01,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","shift_video","shift_audio"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"MiniMaxH3SigmaShift","display_name":"ModelSamplingMiniMaxH3","description":"Set the video/audio flow shifts.","python_module":"comfy_extras.nodes_minimax_h3","category":"model/patch/minimax","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["sigma shift","minimax shift"],"essentials_category":null,"has_intermediate_output":false},"MiniMaxMusic3TextEncode":{"input":{"required":{"clip":["CLIP",{}],"caption":["STRING",{"multiline":true,"dynamicPrompts":true}],"lyrics":["STRING",{"multiline":true,"dynamicPrompts":true}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}],"max_duration":["FLOAT",{"tooltip":"Maximum duration in seconds; the model can end the song earlier.","default":120.0,"min":0.04,"max":360.0,"step":0.04}],"cfg_scale":["FLOAT",{"advanced":true,"default":1.5,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"top_k":["INT",{"advanced":true,"default":50,"min":1,"max":16384}]}},"input_order":{"required":["clip","caption","lyrics","seed","max_duration","cfg_scale","top_k"]},"is_input_list":false,"output":["CONDITIONING","FLOAT"],"output_is_list":[false,false],"output_name":["CONDITIONING","seconds"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"MiniMaxMusic3TextEncode","display_name":"MiniMax Music3 Text Encode","description":"Uses a MiniMax Music3 CLIP model to generate the acoustic conditioning sequence.","python_module":"comfy_extras.nodes_minimax_music","category":"model/conditioning/minimax music","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"MoGeInference":{"input":{"required":{"moge_model":["MOGE_MODEL",{}],"image":["IMAGE",{}],"resolution_level":["INT",{"tooltip":"0 = fastest, 9 = most detail.","default":9,"min":0,"max":9}],"fov_x_degrees":["FLOAT",{"tooltip":"Horizontal field of view of the source camera. Sets the focal length used to unproject the depth map into 3D. 0 = auto-recover from the predicted points.","advanced":true,"default":0.0,"min":0.0,"max":170.0,"step":0.1}],"batch_size":["INT",{"tooltip":"Images per inference call. Lower if you OOM on a long video / image set.","default":4,"min":1,"max":64}],"force_projection":["BOOLEAN",{"advanced":true,"default":true}],"apply_mask":["BOOLEAN",{"tooltip":"Set masked-out (sky / invalid) pixels to inf in points and depth so meshing culls them. Disable to keep the raw predicted geometry everywhere; the mask is still returned separately.","advanced":true,"default":true}],"refine_steps":["INT",{"tooltip":"MoGe-3 only: sparse volumetric refinement passes over the predicted depth. More passes sharpen fine detail and edges at a roughly linear cost. 0 disables refinement. Ignored by MoGe-1 / MoGe-2.","advanced":true,"default":3,"min":0,"max":8}]}},"input_order":{"required":["moge_model","image","resolution_level","fov_x_degrees","batch_size","force_projection","apply_mask","refine_steps"]},"is_input_list":false,"output":["MOGE_GEOMETRY"],"output_is_list":[false],"output_name":["moge_geometry"],"output_tooltips":[null],"output_matchtypes":null,"name":"MoGeInference","display_name":"Run MoGe Inference","description":"Run MoGe on a single image to estimate depth and geometry.","python_module":"comfy_extras.nodes_moge","category":"image/geometry estimation","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["moge","depth","geometry","depth estimation","geometry estimation"],"essentials_category":null,"has_intermediate_output":false},"ModelAttentionBackend":{"input":{"required":{"model":["MODEL",{"tooltip":"The model to patch."}],"attention":["COMBO",{"display_name":"backend","tooltip":"The dense attention backend. Comfy Kitchen attention uses quantized INT8 attention and is available only on Nvidia and AMD GPUs.","default":"pytorch attention","multiselect":false,"options":["pytorch attention"]}]}},"input_order":{"required":["model","attention"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["model"],"output_tooltips":["The model with the selected attention backend."],"output_matchtypes":null,"name":"ModelAttentionBackend","display_name":"Model Attention Backend","description":"Selects the dense attention implementation for the model. When used with Block Sparse Attention, this backend is used whenever sparse attention is inactive or unsupported.","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["comfy kitchen","attention","pytorch","kitchen attention"],"essentials_category":null,"has_intermediate_output":false},"ModelSamplingAuraFlow":{"input":{"required":{"model":["MODEL"],"shift":["FLOAT",{"default":1.73,"min":0.0,"max":100.0,"step":0.01}]},"optional":{"sampling":[["flow","img_to_img_velocity"],{"default":"flow","advanced":true}]}},"input_order":{"required":["model","shift"],"optional":["sampling"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"ModelSamplingAuraFlow","display_name":"ModelSamplingAuraFlow","description":"","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"ModelSamplingFlux":{"input":{"required":{"model":["MODEL"],"max_shift":["FLOAT",{"default":1.15,"min":0.0,"max":100.0,"step":0.01,"advanced":true}],"base_shift":["FLOAT",{"default":0.5,"min":0.0,"max":100.0,"step":0.01,"advanced":true}],"width":["INT",{"default":1024,"min":16,"max":16384,"step":8}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":8}]}},"input_order":{"required":["model","max_shift","base_shift","width","height"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"ModelSamplingFlux","display_name":"ModelSamplingFlux","description":"","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch/flux","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"ModelSamplingSD3":{"input":{"required":{"model":["MODEL"],"shift":["FLOAT",{"default":3.0,"min":0.0,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","shift"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"ModelSamplingSD3","display_name":"ModelSamplingSD3","description":"","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch/stable diffusion","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"PreviewAny":{"input":{"required":{"source":["*",{}]}},"input_order":{"required":["source"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"name":"PreviewAny","display_name":"Preview as Text","description":"Preview any input value as text.","python_module":"comfy_extras.nodes_preview_any","category":"utilities","output_node":true,"has_intermediate_output":false,"search_aliases":["preview","preview text","show output","inspect","debug","print value","show text"]},"PreviewImage":{"input":{"required":{"images":["IMAGE"]},"hidden":{"prompt":"PROMPT","extra_pnginfo":"EXTRA_PNGINFO"}},"input_order":{"required":["images"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"name":"PreviewImage","display_name":"Preview Image","description":"Preview the images without saving them to the ComfyUI output directory.","python_module":"nodes","category":"image","output_node":true,"has_intermediate_output":false,"search_aliases":["preview","preview image","show image","view image","display image","image viewer"],"essentials_category":"Basics"},"PrimitiveBoolean":{"input":{"required":{"value":["BOOLEAN",{}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["BOOLEAN"],"output_is_list":[false],"output_name":["BOOLEAN"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveBoolean","display_name":"Boolean","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"PrimitiveFloat":{"input":{"required":{"value":["FLOAT",{"min":-9223372036854775807,"max":9223372036854775807,"step":0.1}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["FLOAT"],"output_is_list":[false],"output_name":["FLOAT"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveFloat","display_name":"Float","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"PrimitiveInt":{"input":{"required":{"value":["INT",{"min":-9223372036854775807,"max":9223372036854775807,"control_after_generate":"fixed"}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["INT"],"output_is_list":[false],"output_name":["INT"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveInt","display_name":"Int","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"PrimitiveString":{"input":{"required":{"value":["STRING",{"multiline":false}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveString","display_name":"Text","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["text","string","text box","prompt"],"essentials_category":null,"has_intermediate_output":false},"PrimitiveStringMultiline":{"input":{"required":{"value":["STRING",{"multiline":true}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveStringMultiline","display_name":"Text (Multiline)","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["text","string","text box","prompt"],"essentials_category":"Basics","has_intermediate_output":false},"RandomNoise":{"input":{"required":{"noise_seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}]}},"input_order":{"required":["noise_seed"]},"is_input_list":false,"output":["NOISE"],"output_is_list":[false],"output_name":["NOISE"],"output_tooltips":[null],"output_matchtypes":null,"name":"RandomNoise","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/noise","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"ReferenceMultiLatent":{"input":{"required":{"positive":["CONDITIONING",{}],"resolution":["COMFY_DYNAMICCOMBO_V3",{"options":[{"key":"keep resolution","inputs":{"required":{}}},{"key":"resize","inputs":{"required":{"megapixels":["FLOAT",{"tooltip":"Target size of each reference image, preserving aspect ratio. 1.0 is about 1024 x 1024 pixels.","default":1.0,"min":0.01,"max":16.0,"step":0.01}]}}}]}]},"optional":{"negative":["CONDITIONING",{}],"vae":["VAE",{"tooltip":"Reference images are ignored when no VAE is connected."}],"images":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{},"optional":{"image":["IMAGE",{}]}},"prefix":"image","min":1,"max":100}}]}},"input_order":{"required":["positive","resolution"],"optional":["negative","vae","images"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","empty_latent"],"output_tooltips":[null,"Returns the positive output when negative conditioning is not connected.","First reference latent multiplied by 0.0, or a square empty latent at the selected resolution."],"output_matchtypes":null,"name":"ReferenceMultiLatent","display_name":"Set Reference Multi Latent","description":"Encodes reference images and appends their latents to positive and optional negative conditioning for edit models.","python_module":"comfy_extras.nodes_edit_model","category":"model/conditioning","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"RemoveBackground":{"input":{"required":{"bg_removal_model":["BACKGROUND_REMOVAL",{"tooltip":"Background removal model used to generate the mask"}],"image":["IMAGE",{"tooltip":"Input image to remove the background from"}]}},"input_order":{"required":["bg_removal_model","image"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["mask"],"output_tooltips":["Generated foreground mask"],"output_matchtypes":null,"name":"RemoveBackground","display_name":"Remove Background","description":"Generates a foreground mask to remove the background from an image using a background removal model.","python_module":"comfy_extras.nodes_bg_removal","category":"image/background removal","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"RepeatImageBatch":{"input":{"required":{"image":["IMAGE",{}],"amount":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["image","amount"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"RepeatImageBatch","display_name":"Repeat Image Batch","description":"","python_module":"comfy_extras.nodes_images","category":"image/batch","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["duplicate image","clone image"],"essentials_category":null,"has_intermediate_output":false},"ResizeImageMaskNode":{"input":{"required":{"input":["COMFY_MATCHTYPE_V3",{"template":{"template_id":"input_type","allowed_types":"IMAGE,MASK"}}],"resize_type":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"Select how to resize: by exact dimensions, scale factor, matching another image, etc.","options":[{"key":"scale dimensions","inputs":{"required":{"width":["INT",{"tooltip":"Target width in pixels. Set to 0 to auto-calculate from height while preserving aspect ratio.","default":512,"min":0,"max":16384,"step":1}],"height":["INT",{"tooltip":"Target height in pixels. Set to 0 to auto-calculate from width while preserving aspect ratio.","default":512,"min":0,"max":16384,"step":1}],"crop":["COMBO",{"tooltip":"How to handle aspect ratio mismatch: 'disabled' stretches to fit, 'center' crops to maintain aspect ratio.","default":"center","multiselect":false,"options":["disabled","center"]}]}}},{"key":"scale by multiplier","inputs":{"required":{"multiplier":["FLOAT",{"tooltip":"Scale factor (e.g., 2.0 doubles size, 0.5 halves size).","default":1.0,"min":0.01,"max":8.0,"step":0.01}]}}},{"key":"scale longer dimension","inputs":{"required":{"longer_size":["INT",{"tooltip":"The longer edge will be resized to this value. Aspect ratio is preserved.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale shorter dimension","inputs":{"required":{"shorter_size":["INT",{"tooltip":"The shorter edge will be resized to this value. Aspect ratio is preserved.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale width","inputs":{"required":{"width":["INT",{"tooltip":"Target width in pixels. Height auto-adjusts to preserve aspect ratio.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale height","inputs":{"required":{"height":["INT",{"tooltip":"Target height in pixels. Width auto-adjusts to preserve aspect ratio.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale total pixels","inputs":{"required":{"megapixels":["FLOAT",{"tooltip":"Target total megapixels (e.g., 1.0 \u2248 1024\u00d71024). Aspect ratio is preserved.","default":1.0,"min":0.01,"max":16.0,"step":0.01}]}}},{"key":"match size","inputs":{"required":{"match":["IMAGE,MASK",{"tooltip":"Resize input to match the dimensions of this reference image or mask."}],"crop":["COMBO",{"tooltip":"How to handle aspect ratio mismatch: 'disabled' stretches to fit, 'center' crops to maintain aspect ratio.","default":"center","multiselect":false,"options":["disabled","center"]}]}}},{"key":"scale to multiple","inputs":{"required":{"multiple":["INT",{"tooltip":"Resize so width and height are divisible by this number. Useful for latent alignment (e.g., 8 or 64).","default":8,"min":1,"max":16384,"step":1}]}}}]}],"scale_method":["COMBO",{"tooltip":"Interpolation algorithm. 'area' is best for downscaling, 'lanczos' for upscaling, 'nearest-exact' for pixel art.","default":"area","multiselect":false,"options":["nearest-exact","bilinear","area","bicubic","lanczos"]}]}},"input_order":{"required":["input","resize_type","scale_method"]},"is_input_list":false,"output":["COMFY_MATCHTYPE_V3"],"output_is_list":[false],"output_name":["resized"],"output_tooltips":[null],"output_matchtypes":["input_type"],"name":"ResizeImageMaskNode","display_name":"Resize Image/Mask","description":"Resize an image or mask using various scaling methods.","python_module":"comfy_extras.nodes_post_processing","category":"image/transform","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["resize","resize image","resize mask","scale","scale image","scale mask","image resize","change size","dimensions","shrink","enlarge"],"essentials_category":null,"has_intermediate_output":false},"ResolutionSelector":{"input":{"required":{"aspect_ratio":["COMBO",{"tooltip":"The aspect ratio for the output dimensions.","default":"1:1 (Square)","multiselect":false,"options":["1:1 (Square)","2:3 (Portrait Photo)","3:2 (Photo)","3:4 (Portrait Standard)","4:3 (Standard)","9:16 (Portrait Widescreen)","16:9 (Widescreen)","21:9 (Ultrawide)"]}],"megapixels":["FLOAT",{"tooltip":"Target total megapixels. 1.0 MP \u2248 1024x1024 for square.","default":1.0,"min":0.1,"max":16.0,"step":0.1}],"multiple":["INT",{"tooltip":"Nearest multiple of the result to set the selected resolution to.","advanced":true,"default":8,"min":8,"max":128,"step":4}]},"optional":{"preview":["RESOLUTION_PREVIEW",{"tooltip":"Live preview of the calculated output resolution.","socketless":true,"ratio_widget":"aspect_ratio","megapixels_widget":"megapixels","multiple_widget":"multiple"}]}},"input_order":{"required":["aspect_ratio","megapixels","multiple"],"optional":["preview"]},"is_input_list":false,"output":["INT","INT"],"output_is_list":[false,false],"output_name":["width","height"],"output_tooltips":["Calculated width in pixels multiplied by the selected multiple.","Calculated height in pixels multiplied by the selected multiple."],"output_matchtypes":null,"name":"ResolutionSelector","display_name":"Resolution Selector","description":"Calculate width and height from aspect ratio and megapixel target. Useful for setting up Empty Latent Image dimensions.","python_module":"comfy_extras.nodes_resolution","category":"utilities","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SAM3_VideoTrack":{"input":{"required":{"images":["IMAGE",{"display_name":"images","tooltip":"Video frames as batched images"}],"model":["MODEL",{"display_name":"model"}],"detection_threshold":["FLOAT",{"display_name":"detection_threshold","tooltip":"Score threshold for text-prompted detection.","default":0.5,"min":0.0,"max":1.0,"step":0.01}],"max_objects":["INT",{"display_name":"max_objects","tooltip":"Max tracked objects. Initial masks count toward this limit. 0 uses the internal cap of 64.","default":4,"min":0,"max":64}],"detect_interval":["INT",{"display_name":"detect_interval","tooltip":"Run detection every N frames (1=every frame). Higher values save compute.","default":1,"min":1}]},"optional":{"initial_mask":["MASK",{"display_name":"initial_mask","tooltip":"Mask(s) for the first frame to track (one per object)"}],"conditioning":["CONDITIONING",{"display_name":"conditioning","tooltip":"Text conditioning for detecting new objects during tracking"}]}},"input_order":{"required":["images","model","detection_threshold","max_objects","detect_interval"],"optional":["initial_mask","conditioning"]},"is_input_list":false,"output":["SAM3_TRACK_DATA"],"output_is_list":[false],"output_name":["track_data"],"output_tooltips":[null],"output_matchtypes":null,"name":"SAM3_VideoTrack","display_name":"Run SAM3 Video Track","description":"","python_module":"comfy_extras.nodes_sam3","category":"image/detection","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["sam3","video","track","propagate"],"essentials_category":null,"has_intermediate_output":false},"SCAIL2ColoredMask":{"input":{"required":{"driving_track_data":["SAM3_TRACK_DATA",{"tooltip":"SAM3 track of the driving pose video. Will be rendered into the pose_video_mask output."}],"object_indices":["STRING",{"tooltip":"Comma-separated list of person indices to include (e.g. '0,2,3'). Applied to both reference and pose video masks. Empty = all.","default":"","multiline":false}],"sort_by":["COMBO",{"tooltip":"Order in which palette colors are assigned to the tracked objects (applied to both reference and pose video so each identity keeps the same color). Objects that appear in earlier frames always come first; within a frame, left_to_right = leftmost object (by centroid at first appearance) gets the first color, area = biggest object (by mask area at first appearance) gets the first color; none = keep SAM3's order.","default":"left_to_right","multiselect":false,"options":["none","left_to_right","area"]}],"replacement_mode":["BOOLEAN",{"tooltip":"False = Animation Mode (pose_video_mask has black background, reference_image_mask has white background). True = Replacement Mode (pose_video_mask has white background, reference_image_mask has black background).","default":false}]},"optional":{"ref_track_data":["SAM3_TRACK_DATA,MASK",{"display_name":"reference_masks","tooltip":"SAM3 track of the reference image(s) (one identity per object, colored in batch order), or a plain MASK of the reference subject (rendered as a single identity)."}]}},"input_order":{"required":["driving_track_data","object_indices","sort_by","replacement_mode"],"optional":["ref_track_data"]},"is_input_list":false,"output":["IMAGE","IMAGE"],"output_is_list":[false,false],"output_name":["pose_video_mask","reference_image_mask"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"SCAIL2ColoredMask","display_name":"Create SCAIL-2 Colored Mask","description":"","python_module":"comfy_extras.nodes_scail","category":"model/conditioning/wan/scail","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SamplerCustom":{"input":{"required":{"model":["MODEL",{}],"add_noise":["BOOLEAN",{"advanced":true,"default":true}],"noise_seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"sampler":["SAMPLER",{}],"sigmas":["SIGMAS",{}],"latent_image":["LATENT",{}]}},"input_order":{"required":["model","add_noise","noise_seed","cfg","positive","negative","sampler","sigmas","latent_image"]},"is_input_list":false,"output":["LATENT","LATENT"],"output_is_list":[false,false],"output_name":["output","denoised_output"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"SamplerCustom","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/custom","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SamplerCustomAdvanced":{"input":{"required":{"noise":["NOISE",{}],"guider":["GUIDER",{}],"sampler":["SAMPLER",{}],"sigmas":["SIGMAS",{}],"latent_image":["LATENT",{}]}},"input_order":{"required":["noise","guider","sampler","sigmas","latent_image"]},"is_input_list":false,"output":["LATENT","LATENT"],"output_is_list":[false,false],"output_name":["output","denoised_output"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"SamplerCustomAdvanced","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/custom","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SamplerEulerAncestral":{"input":{"required":{"eta":["FLOAT",{"advanced":true,"default":1.0,"min":0.0,"max":100.0,"step":0.01,"round":false}],"s_noise":["FLOAT",{"advanced":true,"default":1.0,"min":0.0,"max":100.0,"step":0.01,"round":false}]}},"input_order":{"required":["eta","s_noise"]},"is_input_list":false,"output":["SAMPLER"],"output_is_list":[false],"output_name":["SAMPLER"],"output_tooltips":[null],"output_matchtypes":null,"name":"SamplerEulerAncestral","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/samplers","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SaveAudioAdvanced":{"input":{"required":{"audio":["AUDIO",{"tooltip":"The audio to save."}],"filename_prefix":["STRING",{"tooltip":"The prefix for the file to save. May include formatting tokens such as %date:yyyy-MM-dd%.","default":"audio/ComfyUI","multiline":false}],"format":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The file format in which to save the audio.","options":[{"key":"flac","inputs":{"required":{}}},{"key":"mp3","inputs":{"required":{"quality":["COMBO",{"default":"V0","multiselect":false,"options":["V0","128k","320k"]}]}}},{"key":"opus","inputs":{"required":{"quality":["COMBO",{"default":"128k","multiselect":false,"options":["64k","96k","128k","192k","320k"]}]}}}]}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["audio","filename_prefix","format"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["audio"],"output_tooltips":[null],"output_matchtypes":null,"name":"SaveAudioAdvanced","display_name":"Save Audio (Advanced)","description":"Saves the input audio to your ComfyUI output directory.","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["save audio","export audio","output audio","write audio","flac","mp3","opus"],"essentials_category":null,"has_intermediate_output":false},"SaveImage":{"input":{"required":{"images":["IMAGE",{"tooltip":"The images to save."}],"filename_prefix":["STRING",{"default":"ComfyUI","tooltip":"The prefix for the file to save. This may include formatting information such as %date:yyyy-MM-dd% or %Empty Latent Image.width% to include values from nodes."}]},"hidden":{"prompt":"PROMPT","extra_pnginfo":"EXTRA_PNGINFO"}},"input_order":{"required":["images","filename_prefix"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"name":"SaveImage","display_name":"Save Image","description":"Saves the input images to your ComfyUI output directory.","python_module":"nodes","category":"image","output_node":true,"has_intermediate_output":false,"search_aliases":["save","save image","export image","output image","write image","download"],"essentials_category":"Basics"},"SaveVideo":{"input":{"required":{"video":["VIDEO",{"tooltip":"The video to save."}],"filename_prefix":["STRING",{"tooltip":"The prefix for the file to save. This may include formatting information such as %date:yyyy-MM-dd% or %Empty Latent Image.width% to include values from nodes.","default":"video/ComfyUI","multiline":false}],"format":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output container. Auto uses MP4 for Auto/H.264 and WebM for AV1. MP4, MKV, and WebM select a specific container.","options":[{"key":"auto","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}},{"key":"mp4","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}},{"key":"mkv","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}},{"key":"webm","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}}]}]},"optional":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","hidden":true,"options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["video","filename_prefix","format"],"optional":["codec"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["video"],"output_tooltips":["The input video, unchanged."],"output_matchtypes":null,"name":"SaveVideo","display_name":"Save Video","description":"Saves the input videos to your ComfyUI output directory.","python_module":"comfy_extras.nodes_video","category":"video","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["export video"],"essentials_category":"Basics","has_intermediate_output":false},"SeedNode":{"input":{"required":{"seed":["INT",{"min":0,"max":9223372036854775807,"control_after_generate":"fixed"}]}},"input_order":{"required":["seed"]},"is_input_list":false,"output":["INT"],"output_is_list":[false],"output_name":["seed"],"output_tooltips":[null],"output_matchtypes":null,"name":"SeedNode","display_name":"Seed","description":"","python_module":"comfy_extras.nodes_seed","category":"utilities","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seed","random"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2Conditioning":{"input":{"required":{"model":["MODEL",{"tooltip":"The SeedVR2 model."}],"vae_conditioning":["LATENT",{"display_name":"latent"}]}},"input_order":{"required":["model","vae_conditioning"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING"],"output_is_list":[false,false],"output_name":["positive","negative"],"output_tooltips":["The positive conditioning for sampling.","The negative conditioning for sampling."],"output_matchtypes":null,"name":"SeedVR2Conditioning","display_name":"Apply SeedVR2 Conditioning","description":"Build SeedVR2 positive/negative conditioning from a VAE latent.","python_module":"comfy_extras.nodes_seedvr","category":"model/conditioning","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","upscale","conditioning"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2PostProcessing":{"input":{"required":{"images":["IMAGE",{"tooltip":"The generated image to process."}],"original_resized_images":["IMAGE",{"tooltip":"The original resized image before pre-processing, used as reference."}],"color_correction_method":["COMBO",{"tooltip":"Method to match the generated image colors to the original image. lab: transfer color in CIELAB space, preserving detail (most faithful). wavelet: transfer low-frequency color, keeping upscaled high-frequency detail. adain: match per-channel mean/std (fastest, global tint). none: skip color transfer (geometry alignment only).","default":"lab","multiselect":false,"options":["lab","wavelet","adain","none"]}]}},"input_order":{"required":["images","original_resized_images","color_correction_method"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"output_tooltips":["The aligned, color-corrected image."],"output_matchtypes":null,"name":"SeedVR2PostProcessing","display_name":"Post-Process SeedVR2 Output","description":"Align the generated image with the original resized image and apply color correction.","python_module":"comfy_extras.nodes_seedvr","category":"image/post-processors","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","upscale","color correction","color match","postprocess"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2Preprocess":{"input":{"required":{"resized_images":["IMAGE",{"tooltip":"The resized image to process."}]}},"input_order":{"required":["resized_images"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"output_tooltips":["The padded image for VAE encoding."],"output_matchtypes":null,"name":"SeedVR2Preprocess","display_name":"Pre-Process SeedVR2 Input","description":"Pad a resized image for SeedVR2 model. Alpha channel is dropped. The node Post-Process SeedVR2 Output re-applies it from the original resized image.","python_module":"comfy_extras.nodes_seedvr","category":"image/pre-processors","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","upscale","video upscale","pad","preprocess"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2TemporalChunk":{"input":{"required":{"latent":["LATENT",{"tooltip":"The VAE-encoded SeedVR2 latent to split."}],"temporal_overlap":["INT",{"tooltip":"Latent frames shared between adjacent chunks and crossfaded at merge; 0 = no overlap.","default":0,"min":0,"max":16384}],"chunking_mode":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"manual = use frames_per_chunk exactly; auto = predict the largest chunk that fits free VRAM.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"manual","inputs":{"required":{"frames_per_chunk":["INT",{"tooltip":"Pixel frames per temporal chunk (4n+1: 1, 5, 9, 13, ...).","default":21,"min":1,"max":16384,"step":4}]}}}]}]}},"input_order":{"required":["latent","temporal_overlap","chunking_mode"]},"is_input_list":false,"output":["LATENT","INT"],"output_is_list":[true,false],"output_name":["latents","temporal_overlap"],"output_tooltips":["The temporal chunks in sequence order.","The effective latent-frame overlap between adjacent chunks, for Merge SeedVR2 Latents."],"output_matchtypes":null,"name":"SeedVR2TemporalChunk","display_name":"Split SeedVR2 Latent","description":"Split a SeedVR2 video latent into overlapping temporal chunks small enough to sample one at a time within VRAM, wiring latents outputs to both Apply SeedVR2 Conditioning and the sampler latent input before recombining with Merge SeedVR2 Latents.","python_module":"comfy_extras.nodes_seedvr","category":"model/latent/seedvr","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","split","chunk","temporal","video upscale","rebatch"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2TemporalMerge":{"input":{"required":{"latents":["LATENT",{"tooltip":"The sampled temporal chunks in sequence order."}],"temporal_overlap":["INT",{"tooltip":"The temporal_overlap output of Split SeedVR2 Latent. 0 = plain concatenation.","default":0,"forceInput":true,"min":0,"max":16384}]}},"input_order":{"required":["latents","temporal_overlap"]},"is_input_list":true,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"output_tooltips":["The recombined full-length latent."],"output_matchtypes":null,"name":"SeedVR2TemporalMerge","display_name":"Merge SeedVR2 Latents","description":"Recombine sampled SeedVR2 latent temporal chunks into one latent, crossfading each overlap with a Hann window sized by the temporal_overlap wired from Split SeedVR2 Latent.","python_module":"comfy_extras.nodes_seedvr","category":"model/latent/seedvr","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","merge","temporal","hann","crossfade"],"essentials_category":null,"has_intermediate_output":false},"SetLatentNoiseMask":{"input":{"required":{"samples":["LATENT"],"mask":["MASK"]}},"input_order":{"required":["samples","mask"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"SetLatentNoiseMask","display_name":"Set Latent Noise Mask","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"SolidMask":{"input":{"required":{"value":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}],"width":["INT",{"default":512,"min":1,"max":16384,"step":1}],"height":["INT",{"default":512,"min":1,"max":16384,"step":1}]}},"input_order":{"required":["value","width","height"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["MASK"],"output_tooltips":[null],"output_matchtypes":null,"name":"SolidMask","display_name":"Create Solid Mask","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"StringConcatenate":{"input":{"required":{"string_a":["STRING",{"multiline":true}],"string_b":["STRING",{"multiline":true}],"delimiter":["STRING",{"default":"","multiline":false}]}},"input_order":{"required":["string_a","string_b","delimiter"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"StringConcatenate","display_name":"Concatenate Text","description":"","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["concatenate","text concat","join text","merge text","combine strings","string concat","append text","combine text"],"essentials_category":null,"has_intermediate_output":false},"StringContains":{"input":{"required":{"string":["STRING",{"multiline":true}],"substring":["STRING",{"multiline":true}],"case_sensitive":["BOOLEAN",{"advanced":true,"default":true}]}},"input_order":{"required":["string","substring","case_sensitive"]},"is_input_list":false,"output":["BOOLEAN"],"output_is_list":[false],"output_name":["contains"],"output_tooltips":[null],"output_matchtypes":null,"name":"StringContains","display_name":"Contains Text","description":"","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["contains","text includes","string includes"],"essentials_category":null,"has_intermediate_output":false},"StringReplace":{"input":{"required":{"string":["STRING",{"multiline":true}],"find":["STRING",{"multiline":true}],"replace":["STRING",{"multiline":true}]}},"input_order":{"required":["string","find","replace"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"StringReplace","display_name":"Replace Text","description":"","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["replace","find and replace","substitute","swap text"],"essentials_category":null,"has_intermediate_output":false},"TextEncodeQwenImageEditPlus":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"multiline":true,"dynamicPrompts":true}]},"optional":{"vae":["VAE",{}],"image1":["IMAGE",{}],"image2":["IMAGE",{}],"image3":["IMAGE",{}]}},"input_order":{"required":["clip","prompt"],"optional":["vae","image1","image2","image3"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"output_tooltips":[null],"output_matchtypes":null,"name":"TextEncodeQwenImageEditPlus","display_name":"Text Encode Qwen Image Edit Plus","description":"","python_module":"comfy_extras.nodes_qwen","category":"model/conditioning/qwen image","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"TextGenerate":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"default":"","multiline":true,"dynamicPrompts":true}],"max_length":["INT",{"default":512,"min":1,"max":32768}],"sampling_mode":["COMFY_DYNAMICCOMBO_V3",{"display_name":"Sampling Mode","options":[{"key":"on","inputs":{"required":{"temperature":["FLOAT",{"default":0.7,"min":0.01,"max":2.0,"step":1e-06}],"top_k":["INT",{"default":64,"min":0,"max":1000}],"top_p":["FLOAT",{"default":0.95,"min":0.0,"max":1.0,"step":0.01}],"min_p":["FLOAT",{"default":0.05,"min":0.0,"max":1.0,"step":0.01}],"repetition_penalty":["FLOAT",{"default":1.05,"min":0.0,"max":5.0,"step":0.01}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615}]},"optional":{"presence_penalty":["FLOAT",{"default":0.0,"min":0.0,"max":5.0,"step":0.01}]}}},{"key":"off","inputs":{"required":{}}}]}]},"optional":{"image":["IMAGE",{}],"video":["IMAGE",{"tooltip":"Video frames as image batch. Assumed to be 24 FPS; subsampled to 1 FPS internally."}],"audio":["AUDIO",{}],"thinking":["BOOLEAN",{"tooltip":"Operate in thinking mode if the model supports it.","default":false}],"use_default_template":["BOOLEAN",{"tooltip":"Use the built in system prompt/template if the model has one.","advanced":true,"default":true}],"mtp":["COMBO",{"tooltip":"Speculative decoding with the checkpoint's multi-token-prediction head. No effect without MTP weights. auto adapts the draft depth; 2-5 pins it. Sampled output stays correctly distributed but differs from non-MTP output for the same seed.","default":"auto","multiselect":false,"options":["auto","off","2","3","4","5"]}],"system_prompt":["STRING",{"tooltip":"Replaces the system prompt in the model's chat template. Ignored when the default template is not used.","forceInput":true,"multiline":false}]}},"input_order":{"required":["clip","prompt","max_length","sampling_mode"],"optional":["image","video","audio","thinking","use_default_template","mtp","system_prompt"]},"is_input_list":false,"output":["STRING","STRING"],"output_is_list":[false,false],"output_name":["generated_text","thinking"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"TextGenerate","display_name":"Generate Text","description":"","python_module":"comfy_extras.nodes_textgen","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["LLM","gemma"],"essentials_category":null,"has_intermediate_output":false},"TextGenerateLTX2Prompt":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"default":"","multiline":true,"dynamicPrompts":true}],"max_length":["INT",{"default":512,"min":1,"max":32768}],"sampling_mode":["COMFY_DYNAMICCOMBO_V3",{"display_name":"Sampling Mode","options":[{"key":"on","inputs":{"required":{"temperature":["FLOAT",{"default":0.7,"min":0.01,"max":2.0,"step":1e-06}],"top_k":["INT",{"default":64,"min":0,"max":1000}],"top_p":["FLOAT",{"default":0.95,"min":0.0,"max":1.0,"step":0.01}],"min_p":["FLOAT",{"default":0.05,"min":0.0,"max":1.0,"step":0.01}],"repetition_penalty":["FLOAT",{"default":1.05,"min":0.0,"max":5.0,"step":0.01}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615}]},"optional":{"presence_penalty":["FLOAT",{"default":0.0,"min":0.0,"max":5.0,"step":0.01}]}}},{"key":"off","inputs":{"required":{}}}]}]},"optional":{"image":["IMAGE",{}],"video":["IMAGE",{"tooltip":"Video frames as image batch. Assumed to be 24 FPS; subsampled to 1 FPS internally."}],"audio":["AUDIO",{}],"thinking":["BOOLEAN",{"tooltip":"Operate in thinking mode if the model supports it.","default":false}],"use_default_template":["BOOLEAN",{"tooltip":"Use the built in system prompt/template if the model has one.","advanced":true,"default":true}],"mtp":["COMBO",{"tooltip":"Speculative decoding with the checkpoint's multi-token-prediction head. No effect without MTP weights. auto adapts the draft depth; 2-5 pins it. Sampled output stays correctly distributed but differs from non-MTP output for the same seed.","default":"auto","multiselect":false,"options":["auto","off","2","3","4","5"]}],"system_prompt":["STRING",{"tooltip":"Replaces the system prompt in the model's chat template. Ignored when the default template is not used.","forceInput":true,"multiline":false}]}},"input_order":{"required":["clip","prompt","max_length","sampling_mode"],"optional":["image","video","audio","thinking","use_default_template","mtp","system_prompt"]},"is_input_list":false,"output":["STRING","STRING"],"output_is_list":[false,false],"output_name":["generated_text","thinking"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"TextGenerateLTX2Prompt","display_name":"Generate LTX2 Prompt","description":"","python_module":"comfy_extras.nodes_textgen","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["prompt enhance","LLM","gemma"],"essentials_category":null,"has_intermediate_output":false},"TrimAudioDuration":{"input":{"required":{"audio":["AUDIO",{}],"start_index":["FLOAT",{"tooltip":"Start time in seconds, can be negative to count from the end (supports sub-seconds).","default":0.0,"min":-18446744073709551615,"max":18446744073709551615,"step":0.01}],"duration":["FLOAT",{"tooltip":"Duration in seconds","default":60.0,"min":0.0,"step":0.01}]}},"input_order":{"required":["audio","start_index","duration"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"TrimAudioDuration","display_name":"Trim Audio Duration","description":"Trim audio tensor into chosen time range.","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["cut audio","audio clip","shorten audio"],"essentials_category":null,"has_intermediate_output":false},"TrimVideoLatent":{"input":{"required":{"samples":["LATENT",{}],"trim_amount":["INT",{"default":0,"min":0,"max":99999}]}},"input_order":{"required":["samples","trim_amount"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"TrimVideoLatent","display_name":"Trim Video Latent","description":"","python_module":"comfy_extras.nodes_wan","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"UNETLoader":{"input":{"required":{"unet_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]],"weight_dtype":[["default","fp8_e4m3fn","fp8_e4m3fn_fast","fp8_e5m2"],{"advanced":true}]}},"input_order":{"required":["unet_name","weight_dtype"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"UNETLoader","display_name":"Load Diffusion Model","description":"","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"VAEDecode":{"input":{"required":{"samples":["LATENT",{"tooltip":"The latent to be decoded."}],"vae":["VAE",{"tooltip":"The VAE model used for decoding the latent."}]}},"input_order":{"required":["samples","vae"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"VAEDecode","display_name":"VAE Decode","description":"Decodes latent images back into pixel space images.","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"output_tooltips":["The decoded image."],"search_aliases":["decode","decode latent","latent to image","render latent"]},"VAEDecodeAudio":{"input":{"required":{"samples":["LATENT",{}],"vae":["VAE",{}]}},"input_order":{"required":["samples","vae"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"VAEDecodeAudio","display_name":"VAE Decode Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["latent to audio"],"essentials_category":null,"has_intermediate_output":false},"VAEDecodeAudioTiled":{"input":{"required":{"samples":["LATENT",{}],"vae":["VAE",{}],"tile_size":["INT",{"default":512,"min":32,"max":8192,"step":8}],"overlap":["INT",{"default":64,"min":0,"max":1024,"step":8}]}},"input_order":{"required":["samples","vae","tile_size","overlap"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"VAEDecodeAudioTiled","display_name":"VAE Decode Audio (Tiled)","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["latent to audio"],"essentials_category":null,"has_intermediate_output":false},"VAEDecodeTiled":{"input":{"required":{"samples":["LATENT"],"vae":["VAE"],"tile_size":["INT",{"default":512,"min":64,"max":4096,"step":32,"advanced":true}],"overlap":["INT",{"default":64,"min":0,"max":4096,"step":32,"advanced":true}],"temporal_size":["INT",{"default":64,"min":8,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to decode at a time.","advanced":true}],"temporal_overlap":["INT",{"default":8,"min":4,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to overlap.","advanced":true}]}},"input_order":{"required":["samples","vae","tile_size","overlap","temporal_size","temporal_overlap"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"VAEDecodeTiled","display_name":"VAE Decode (Tiled)","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"VAEEncode":{"input":{"required":{"pixels":["IMAGE"],"vae":["VAE"]}},"input_order":{"required":["pixels","vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"VAEEncode","display_name":"VAE Encode","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":["encode","encode image","image to latent"]},"VAEEncodeAudio":{"input":{"required":{"audio":["AUDIO",{}],"vae":["VAE",{}]}},"input_order":{"required":["audio","vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"VAEEncodeAudio","display_name":"VAE Encode Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["audio to latent"],"essentials_category":null,"has_intermediate_output":false},"VAEEncodeTiled":{"input":{"required":{"pixels":["IMAGE"],"vae":["VAE"],"tile_size":["INT",{"default":512,"min":64,"max":4096,"step":64,"advanced":true}],"overlap":["INT",{"default":64,"min":0,"max":4096,"step":32,"advanced":true}],"temporal_size":["INT",{"default":64,"min":8,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to encode at a time.","advanced":true}],"temporal_overlap":["INT",{"default":8,"min":4,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to overlap.","advanced":true}]}},"input_order":{"required":["pixels","vae","tile_size","overlap","temporal_size","temporal_overlap"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"VAEEncodeTiled","display_name":"VAE Encode (Tiled)","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"VAELoader":{"input":{"required":{"vae_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors","pixel_space"]]}},"input_order":{"required":["vae_name"]},"is_input_list":false,"output":["VAE"],"output_is_list":[false],"output_name":["VAE"],"name":"VAELoader","display_name":"Load VAE","description":"","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"Video Slice":{"input":{"required":{"video":["VIDEO",{}],"start_time":["FLOAT",{"tooltip":"Start time in seconds","default":0.0,"min":-100000.0,"max":100000.0,"step":0.001}],"duration":["FLOAT",{"tooltip":"Duration in seconds, or 0 for unlimited duration","default":0.0,"min":0.0,"step":0.001}],"strict_duration":["BOOLEAN",{"tooltip":"If True, when the specified duration is not possible, an error will be raised.","default":false}]}},"input_order":{"required":["video","start_time","duration","strict_duration"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["VIDEO"],"output_tooltips":[null],"output_matchtypes":null,"name":"Video Slice","display_name":"Trim Video","description":"","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["trim video duration","skip first frames","frame load cap","start time"],"essentials_category":"Video Tools","has_intermediate_output":false},"WanAnimate2Cache":{"input":{"required":{"model":["MODEL",{}],"device":["COMBO",{"tooltip":"Where to keep the cache. cpu (RAM) is the safe choice, the cache will not fit in VRAM alongside the model at typical sizes. gpu (VRAM) can be faster if it fits.","default":"cpu","multiselect":false,"options":["cpu","gpu"]}],"dtype":["COMBO",{"tooltip":"Storage precision. default stores the activations in the model's compute dtype. int8 halves the cache, int4 quarters it, convrot is used to retain accuracy.","default":"default","multiselect":false,"options":["default","int8","int4"]}]}},"input_order":{"required":["model","device","dtype"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"WanAnimate2Cache","display_name":null,"description":"Caches the pose-video's per-block activations so it runs once instead of on every sampling step. Roughly halves generation time Tradeoff is ~12.5 GB of system RAM at 480x832/81 frames in bf16 (scales with resolution and length). With context windows each window is cached separately, so RAM scales with the window count; use the static_standard schedule, as uniform schedules shift the windows every step and nothing ever recurs to hit the cache.","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan/animate","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanAnimate2ToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"tooltip":"Output video width in pixels.","default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"tooltip":"Output video height in pixels.","default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"tooltip":"Number of frames to generate.","default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"tooltip":"Number of videos to generate simultaneously.","default":1,"min":1,"max":4096}],"video_frame_offset":["INT",{"tooltip":"Frames to seek into the pose video. Connect to the video_frame_offset output of the previous node when extending.","default":0,"min":0,"max":16384,"step":1}],"pose_strength":["FLOAT",{"tooltip":"Scales the pose video's influence on the motion. 1.0 is the trained behavior; below weakens adherence, above amplifies. 0.0 mutes it but does not fully remove it.","default":1.0,"min":0.0,"max":10.0,"step":0.01}],"pose_start_percent":["FLOAT",{"tooltip":"Sampling percent at which the pose influence starts. Outside the window the pose branch is skipped entirely, which also speeds those steps up.","default":0.0,"min":0.0,"max":1.0,"step":0.01}],"pose_end_percent":["FLOAT",{"tooltip":"Sampling percent at which the pose influence ends. Motion is mostly established early, so e.g. 0.7 can loosen fine detail while keeping the choreography.","default":1.0,"min":0.0,"max":1.0,"step":0.01}],"reference_image_strength":["FLOAT",{"tooltip":"Scales how strongly generated frames attend to the reference image's latent frame. Below 1.0 loosens identity/appearance adherence (e.g. to let the prompt restyle), above tightens it against drift.","default":1.0,"min":0.0,"max":10.0,"step":0.01}]},"optional":{"reference_image":["IMAGE",{"tooltip":"The character to animate."}],"pose_video":["IMAGE",{"tooltip":"The video whose motion is transferred to the reference character."}],"clip_vision_output":["CLIP_VISION_OUTPUT",{"tooltip":"CLIP vision of the reference image."}],"positive_pose":["CONDITIONING",{"tooltip":"Prompt for the pose-video branch, describing the motion rather than the character. Defaults to positive. Used for both the cond and uncond passes."}],"clip_vision_output_pose":["CLIP_VISION_OUTPUT",{"tooltip":"CLIP vision of the pose video's first frame. Defaults to clip_vision_output."}],"continue_motion":["IMAGE",{"tooltip":"Previous motion sequence to continue from for temporal consistency."}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size","video_frame_offset","pose_strength","pose_start_percent","pose_end_percent","reference_image_strength"],"optional":["reference_image","pose_video","clip_vision_output","positive_pose","clip_vision_output_pose","continue_motion"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT","INT","INT","INT"],"output_is_list":[false,false,false,false,false,false],"output_name":["positive","negative","latent","trim_latent","trim_image","video_frame_offset"],"output_tooltips":[null,null,null,"Number of latent frames that should be trimmed before decoding.","Number of overlapping image frames when extending a video.","Frames to seek into the pose video."],"output_matchtypes":null,"name":"WanAnimate2ToVideo","display_name":null,"description":"Animate a character in a reference image using a video, effectively replicating the facial animation, body motion and hands gesture from the video.","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan/animate","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanCameraEmbedding":{"input":{"required":{"camera_pose":["COMBO",{"default":"Static","multiselect":false,"options":["Static","Pan Up","Pan Down","Pan Left","Pan Right","Zoom In","Zoom Out","Anti Clockwise (ACW)","ClockWise (CW)"]}],"width":["INT",{"default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}]},"optional":{"speed":["FLOAT",{"default":1.0,"min":0,"max":10.0,"step":0.1}],"fx":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":1e-09}],"fy":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":1e-09}],"cx":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":0.01}],"cy":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":0.01}]}},"input_order":{"required":["camera_pose","width","height","length"],"optional":["speed","fx","fy","cx","cy"]},"is_input_list":false,"output":["WAN_CAMERA_EMBEDDING","INT","INT","INT"],"output_is_list":[false,false,false,false],"output_name":["camera_embedding","width","height","length"],"output_tooltips":[null,null,null,null],"output_matchtypes":null,"name":"WanCameraEmbedding","display_name":null,"description":"","python_module":"comfy_extras.nodes_camera_trajectory","category":"model/conditioning/wan/camera","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanCameraImageToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]},"optional":{"clip_vision_output":["CLIP_VISION_OUTPUT",{}],"start_image":["IMAGE",{}],"camera_conditions":["WAN_CAMERA_EMBEDDING",{}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size"],"optional":["clip_vision_output","start_image","camera_conditions"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"WanCameraImageToVideo","display_name":null,"description":"","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan/camera","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanImageToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]},"optional":{"clip_vision_output":["CLIP_VISION_OUTPUT",{}],"start_image":["IMAGE",{}],"ref_pad_image":["IMAGE",{"tooltip":"Fills the padding frames of the image conditioning with this image instead of gray, anchoring identity without pinning frames (SVI-style anti-drift padding, used by models such as ID-V2V)."}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size"],"optional":["clip_vision_output","start_image","ref_pad_image"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"WanImageToVideo","display_name":null,"description":"","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanSCAILToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"default":512,"min":32,"max":16384,"step":32}],"height":["INT",{"default":896,"min":32,"max":16384,"step":32}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"default":1,"min":1,"max":4096}],"pose_strength":["FLOAT",{"tooltip":"Strength of the pose latent.","default":1.0,"min":0.0,"max":10.0,"step":0.01}],"pose_start":["FLOAT",{"tooltip":"Start step of the pose conditioning.","default":0.0,"min":0.0,"max":1.0,"step":0.01}],"pose_end":["FLOAT",{"tooltip":"End step of the pose conditioning.","default":1.0,"min":0.0,"max":1.0,"step":0.01}],"video_frame_offset":["INT",{"tooltip":"Cumulative output frame this chunk begins at. Wire from the previous chunk's video_frame_offset output.","default":0,"min":0,"max":16384,"step":1}],"previous_frame_count":["INT",{"tooltip":"Tail frames of previous_frames to anchor. SCAIL-2 trained at 5 (81-frame chunks, 76-frame step).","default":5,"min":1,"max":16384,"step":4}]},"optional":{"pose_video":["IMAGE",{"tooltip":"Video used for pose conditioning. Will be downscaled to half the resolution of the main video."}],"pose_video_mask":["IMAGE",{"tooltip":"SCAIL-2 only. Colored per-identity SAM3 mask video at the same resolution as pose_video."}],"replacement_mode":["BOOLEAN",{"tooltip":"SCAIL-2 only. False = Animation Mode (pose_video_mask should have black background). True = Replacement Mode (pose_video_mask should have white background).","default":false}],"reference_image":["IMAGE",{"tooltip":"Reference image. The first image is the primary reference (composite all identities onto it). SCAIL-2: extra batch images are used as additional views (back view, close-up, occluded background), each needing a matching reference_image_mask in that identity's color."}],"reference_image_mask":["IMAGE",{"tooltip":"SCAIL-2 only. Colored reference mask, batch matching reference_image (first = primary reference mask, rest = identity masks for the additional reference_image)."}],"clip_vision_output":["CLIP_VISION_OUTPUT",{"tooltip":"CLIP vision features for conditioning. Model is trained with stretch resize to aspect ratio."}],"previous_frames":["IMAGE",{"tooltip":"SCAIL-2 only. Full decoded output of the previous chunk. Only the last previous_frame_count are used as the extension anchor."}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size","pose_strength","pose_start","pose_end","video_frame_offset","previous_frame_count"],"optional":["pose_video","pose_video_mask","replacement_mode","reference_image","reference_image_mask","clip_vision_output","previous_frames"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT","INT"],"output_is_list":[false,false,false,false],"output_name":["positive","negative","latent","video_frame_offset"],"output_tooltips":[null,null,"Empty latent of the generation size.","Adjusted offset + length. Wire into the next chunk."],"output_matchtypes":null,"name":"WanSCAILToVideo","display_name":null,"description":"","python_module":"comfy_extras.nodes_scail","category":"model/conditioning/wan/scail","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false}}
+{"AudioConcat":{"input":{"required":{"audio1":["AUDIO",{}],"audio2":["AUDIO",{}],"direction":["COMBO",{"tooltip":"Whether to append audio2 after or before audio1.","default":"after","multiselect":false,"options":["after","before"]}]}},"input_order":{"required":["audio1","audio2","direction"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"AudioConcat","display_name":"Concatenate Audio","description":"Concatenates the audio1 to audio2 in the specified direction.","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["join audio","combine audio","append audio"],"essentials_category":null,"has_intermediate_output":false},"AudioMerge":{"input":{"required":{"audio1":["AUDIO",{}],"audio2":["AUDIO",{}],"merge_method":["COMBO",{"tooltip":"The method used to combine the audio waveforms.","multiselect":false,"options":["add","mean","subtract","multiply"]}]}},"input_order":{"required":["audio1","audio2","merge_method"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"AudioMerge","display_name":"Merge Audio","description":"Combine two audio tracks by overlaying their waveforms.","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["mix audio","overlay audio","layer audio"],"essentials_category":null,"has_intermediate_output":false},"BasicGuider":{"input":{"required":{"model":["MODEL",{}],"conditioning":["CONDITIONING",{}]}},"input_order":{"required":["model","conditioning"]},"is_input_list":false,"output":["GUIDER"],"output_is_list":[false],"output_name":["GUIDER"],"output_tooltips":[null],"output_matchtypes":null,"name":"BasicGuider","display_name":"Basic Guider","description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/guiders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"BasicScheduler":{"input":{"required":{"model":["MODEL",{}],"scheduler":["COMBO",{"multiselect":false,"options":["simple","sgm_uniform","karras","exponential","ddim_uniform","beta","normal","linear_quadratic","kl_optimal"]}],"steps":["INT",{"default":20,"min":1,"max":10000}],"denoise":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}]}},"input_order":{"required":["model","scheduler","steps","denoise"]},"is_input_list":false,"output":["SIGMAS"],"output_is_list":[false],"output_name":["SIGMAS"],"output_tooltips":[null],"output_matchtypes":null,"name":"BasicScheduler","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/schedulers","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"BatchImagesNode":{"input":{"required":{"images":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"image":["IMAGE",{}]}},"prefix":"image","min":1,"max":50}}]}},"input_order":{"required":["images"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"BatchImagesNode","display_name":"Batch Images","description":"","python_module":"comfy_extras.nodes_post_processing","category":"image/batch","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["batch","image batch","batch images","combine images","merge images","stack images"],"essentials_category":"Image Tools","has_intermediate_output":false},"CFGGuider":{"input":{"required":{"model":["MODEL",{}],"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}]}},"input_order":{"required":["model","positive","negative","cfg"]},"is_input_list":false,"output":["GUIDER"],"output_is_list":[false],"output_name":["GUIDER"],"output_tooltips":[null],"output_matchtypes":null,"name":"CFGGuider","display_name":"CFG Guider","description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/guiders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CFGNorm":{"input":{"required":{"model":["MODEL",{}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":100.0,"step":0.01}]},"optional":{"pre_cfg":["BOOLEAN",{"tooltip":"If true, rescale the combined noise BEFORE the sampler's CFG combine, without clamping (can amplify). Matches the norm-scaled CFG used by models like Lens. Default false keeps the original post-CFG x0-space attenuate-only behavior.","default":false}]}},"input_order":{"required":["model","strength"],"optional":["pre_cfg"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["patched_model"],"output_tooltips":[null],"output_matchtypes":null,"name":"CFGNorm","display_name":null,"description":"","python_module":"comfy_extras.nodes_cfg","category":"advanced/guidance","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CLIPLoader":{"input":{"required":{"clip_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]],"type":[["stable_diffusion","stable_cascade","sd3","stable_audio","mochi","ltxv","pixart","cosmos","lumina2","wan","hidream","chroma","ace","omnigen2","qwen_image","hunyuan_image","flux2","ovis","longcat_image","cogvideox","lens","pixeldit","ideogram4","boogu","krea2","joyimage","mage","minimax","yue2","flux3","moss_image"]]},"optional":{"device":[["default","cpu"],{"advanced":true}]}},"input_order":{"required":["clip_name","type"],"optional":["device"]},"is_input_list":false,"output":["CLIP"],"output_is_list":[false],"output_name":["CLIP"],"name":"CLIPLoader","display_name":"Load CLIP","description":"Recipes:\nsd: clip-l\nstable cascade: clip-g\nsd3: t5 xxl / clip-g / clip-l\nstable audio: t5 base\nmochi: t5 xxl\ncogvideox: t5 xxl (226-token padding)\ncosmos: old t5 xxl\nlumina2: gemma 2 2B\nwan: umt5 xxl\nhidream: llama-3.1 (Recommend) or t5\nomnigen2: qwen vl 2.5 3B\njoyimage: qwen3-vl 8B\nlens: gpt-oss-20b\npixeldit: gemma 2 2B elm\nminimax: MiniMax H3 Qwen3-VL or Music3 Qwen/RVQ\nflux3: qwen3-vl 4B","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"CLIPTextEncode":{"input":{"required":{"text":["STRING",{"multiline":true,"dynamicPrompts":true,"tooltip":"The text to be encoded."}],"clip":["CLIP",{"tooltip":"The CLIP model used for encoding the text."}]}},"input_order":{"required":["text","clip"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"name":"CLIPTextEncode","display_name":"CLIP Text Encode (Prompt)","description":"Encodes a text prompt using a CLIP model into an embedding that can be used to guide the diffusion model towards generating specific images.","python_module":"nodes","category":"model/conditioning","output_node":false,"has_intermediate_output":false,"output_tooltips":["A conditioning containing the embedded text used to guide the diffusion model."],"search_aliases":["text","prompt","text prompt","positive prompt","negative prompt","encode text","text encoder","encode prompt"]},"CLIPVisionEncode":{"input":{"required":{"clip_vision":["CLIP_VISION"],"image":["IMAGE"],"crop":[["center","none"]]}},"input_order":{"required":["clip_vision","image","crop"]},"is_input_list":false,"output":["CLIP_VISION_OUTPUT"],"output_is_list":[false],"output_name":["CLIP_VISION_OUTPUT"],"name":"CLIPVisionEncode","display_name":"CLIP Vision Encode","description":"","python_module":"nodes","category":"model/conditioning","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"CLIPVisionLoader":{"input":{"required":{"clip_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]]}},"input_order":{"required":["clip_name"]},"is_input_list":false,"output":["CLIP_VISION"],"output_is_list":[false],"output_name":["CLIP_VISION"],"name":"CLIPVisionLoader","display_name":"Load CLIP Vision","description":"","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"Canny":{"input":{"required":{"image":["IMAGE",{}],"low_threshold":["FLOAT",{"default":0.4,"min":0.01,"max":0.99,"step":0.01}],"high_threshold":["FLOAT",{"default":0.8,"min":0.01,"max":0.99,"step":0.01}]}},"input_order":{"required":["image","low_threshold","high_threshold"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"Canny","display_name":"Detect Edges (Canny)","description":"","python_module":"comfy_extras.nodes_canny","category":"image/filters","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["edge detection","outline","contour detection","line art"],"essentials_category":"Image Tools","has_intermediate_output":false},"CheckpointLoaderSimple":{"input":{"required":{"ckpt_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"],{"tooltip":"The name of the checkpoint (model) to load."}]}},"input_order":{"required":["ckpt_name"]},"is_input_list":false,"output":["MODEL","CLIP","VAE"],"output_is_list":[false,false,false],"output_name":["MODEL","CLIP","VAE"],"name":"CheckpointLoaderSimple","display_name":"Load Checkpoint","description":"Loads a diffusion model checkpoint, diffusion models are used to denoise latents.","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"output_tooltips":["The model used for denoising latents.","The CLIP model used for encoding text prompts.","The VAE model used for encoding and decoding images to and from latent space."],"search_aliases":["load model","checkpoint","model loader","load checkpoint","ckpt","model"]},"ColorTransfer":{"input":{"required":{"image_target":["IMAGE",{"tooltip":"Image(s) to apply the color transform to."}],"image_ref":["IMAGE",{"tooltip":"Reference image(s) to match colors to."}],"method":["COMBO",{"multiselect":false,"options":["reinhard_lab","mkl_lab","histogram"]}],"source_stats":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"per_frame: each frame matched to image_ref individually. uniform: pool stats across all source frames as baseline, match to image_ref. target_frame: use one chosen frame as the baseline for the transform to image_ref, applied uniformly to all frames (preserves relative differences)","options":[{"key":"per_frame","inputs":{"required":{}}},{"key":"uniform","inputs":{"required":{}}},{"key":"target_frame","inputs":{"required":{"target_index":["INT",{"tooltip":"Frame index used as the source baseline for computing the transform to image_ref","default":0,"min":0,"max":10000}]}}}]}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":10.0,"step":0.01}]}},"input_order":{"required":["image_target","image_ref","method","source_stats","strength"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["image"],"output_tooltips":[null],"output_matchtypes":null,"name":"ColorTransfer","display_name":"Transfer Color","description":"Match the colors of one image to another using various algorithms.","python_module":"comfy_extras.nodes_post_processing","category":"image/filters","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["color match","color grading","color correction","match colors","color transform","mkl","reinhard","histogram"],"essentials_category":null,"has_intermediate_output":false},"ComfyMathExpression":{"input":{"required":{"expression":["STRING",{"default":"a + b","multiline":true}],"values":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"value":["FLOAT,INT,BOOLEAN",{}]}},"names":["a","b","c","d","e","f","g","h","i","j","k","l","m","n","o","p","q","r","s","t","u","v","w","x","y","z"],"min":1}}]}},"input_order":{"required":["expression","values"]},"is_input_list":false,"output":["FLOAT","INT","BOOLEAN"],"output_is_list":[false,false,false],"output_name":["FLOAT","INT","BOOL"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"ComfyMathExpression","display_name":"Math Expression","description":"","python_module":"comfy_extras.nodes_math","category":"utilities","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["math","expression","formula","calculate","calculator","eval"],"essentials_category":null,"has_intermediate_output":false},"ComfyNotNode":{"input":{"required":{"value":["*",{}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["BOOLEAN"],"output_is_list":[false],"output_name":["BOOLEAN"],"output_tooltips":[null],"output_matchtypes":null,"name":"ComfyNotNode","display_name":"Not","description":"Logical NOT operation. Returns true if the value is falsy. Uses Python's rules for truthiness.","python_module":"comfy_extras.nodes_logic","category":"utilities/logic","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["invert","toggle","negate","flip boolean"],"essentials_category":null,"has_intermediate_output":false},"ComfySwitchNode":{"input":{"required":{"switch":["BOOLEAN",{}]},"optional":{"on_false":["COMFY_MATCHTYPE_V3",{"lazy":true,"template":{"template_id":"switch","allowed_types":"*"}}],"on_true":["COMFY_MATCHTYPE_V3",{"lazy":true,"template":{"template_id":"switch","allowed_types":"*"}}]}},"input_order":{"required":["switch"],"optional":["on_false","on_true"]},"is_input_list":false,"output":["COMFY_MATCHTYPE_V3"],"output_is_list":[false],"output_name":["output"],"output_tooltips":[null],"output_matchtypes":["switch"],"name":"ComfySwitchNode","display_name":"If/Else Switch","description":"","python_module":"comfy_extras.nodes_logic","category":"utilities/logic","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["if","then","switch","conditional","branch"],"essentials_category":null,"has_intermediate_output":false},"ConditioningZeroOut":{"input":{"required":{"conditioning":["CONDITIONING"]}},"input_order":{"required":["conditioning"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"name":"ConditioningZeroOut","display_name":"Conditioning Zero Out","description":"","python_module":"nodes","category":"model/conditioning/transform","output_node":false,"has_intermediate_output":false,"search_aliases":["null conditioning","clear conditioning"]},"ContextWindowsManual":{"input":{"required":{"model":["MODEL",{"tooltip":"The model to apply context windows to during sampling."}],"context_length":["INT",{"tooltip":"The length of the context window.","default":16,"min":1}],"context_overlap":["INT",{"tooltip":"The overlap of the context window.","default":4,"min":0}],"context_schedule":["COMBO",{"tooltip":"Step-dependent scheduling algorithm for context windows.","default":"standard_static","multiselect":false,"options":["standard_static","standard_uniform","looped_uniform","batched"]}],"context_stride":["INT",{"tooltip":"The stride of the context window; only applicable to uniform schedules.","default":1,"min":1}],"closed_loop":["BOOLEAN",{"tooltip":"Whether to close the context window loop; only applicable to looped schedules.","default":false}],"fuse_method":["COMBO",{"tooltip":"The method to use to fuse the context windows.","default":"pyramid","multiselect":false,"options":["pyramid","relative","flat","overlap-linear"]}],"dim":["INT",{"tooltip":"The dimension to apply the context windows to.","default":0,"min":0,"max":5}],"freenoise":["BOOLEAN",{"tooltip":"Whether to apply FreeNoise noise shuffling, improves window blending.","default":false}],"cond_retain_index_list":["STRING",{"tooltip":"List of latent indices to retain in the conditioning tensors for each window. For concat-style I2V models (e.g. Wan I2V, HunyuanVideo I2V, Cosmos I2V, SVD) the encoded start image lives in the c_concat conditioning channels; setting this to '0' will retain that start image content at sub-pos 0 of every window.","default":"","multiline":false}],"split_conds_to_windows":["BOOLEAN",{"tooltip":"Whether to split multiple conditionings (created by ConditionCombine) to each window based on region index.","default":false}],"latent_retain_index_list":["STRING",{"tooltip":"List of latent indices to retain in the noise latent itself for each window. Use for workflows where reference content (e.g. a start image) lives directly in the noise latent rather than in separate conditioning channels (e.g. inplace-style I2V like LTXV, AnimateDiff). Independent of cond_retain_index_list.","default":"","multiline":false}],"causal_window_fix":["BOOLEAN",{"tooltip":"Whether to add a causal fix frame to non-0-indexed context windows.","default":true}]}},"input_order":{"required":["model","context_length","context_overlap","context_schedule","context_stride","closed_loop","fuse_method","dim","freenoise","cond_retain_index_list","split_conds_to_windows","latent_retain_index_list","causal_window_fix"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"output_tooltips":["The model with context windows applied during sampling."],"output_matchtypes":null,"name":"ContextWindowsManual","display_name":"Context Windows (Manual)","description":"Manually set context windows.","python_module":"comfy_extras.nodes_context_windows","category":"model/patch","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CreateBoundingBoxes":{"input":{"required":{"width":["INT",{"tooltip":"Width of the canvas and the pixel grid for the bounding boxes.","default":1024,"min":64,"max":16384,"step":16}],"height":["INT",{"tooltip":"Height of the canvas and the pixel grid for the bounding boxes.","default":1024,"min":64,"max":16384,"step":16}],"editor_state":["BOUNDING_BOXES",{"tooltip":"Draw bounding boxes and set each box type, text, description, color palette. Start with background element first and foreground last.","default":[],"socketless":false}]},"optional":{"background":["IMAGE",{"tooltip":"Optional image used as background in the canvas and preview."}],"bboxes":["BOUNDING_BOX,ARRAY,STRING",{"tooltip":"Bounding boxes, elements, or a JSON string to initialize the canvas. A new upstream value initializes the canvas; edits made on the canvas take priority and are kept until the upstream value changes again."}],"last_incoming":["BOUNDING_BOXES",{"tooltip":"Internal state managed by the canvas: the upstream bboxes value that last initialized it. Leave empty to re-initialize the canvas from the bboxes input on the next run.","default":[],"socketless":true}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["width","height","editor_state"],"optional":["background","bboxes","last_incoming"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["IMAGE","BOUNDING_BOX","ARRAY"],"output_is_list":[false,false,false],"output_name":["preview","bboxes","elements"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"CreateBoundingBoxes","display_name":"Create Bounding Boxes","description":"Draw bounding boxes in a canvas. Outputs Ideogram prompt elements, pixel-space bounding boxes, and a preview image.","python_module":"comfy_extras.nodes_bounding_boxes","category":"utilities","output_node":true,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"CreateVideo":{"input":{"required":{"images":["IMAGE",{"tooltip":"The images to create a video from."}],"fps":["FLOAT",{"default":30.0,"min":1.0,"max":120.0,"step":1.0}]},"optional":{"audio":["AUDIO",{"tooltip":"The audio to add to the video."}],"bit_depth":["COMBO",{"tooltip":"Auto uses 8-bit for sRGB and 10-bit for HDR. Explicit 8-bit and 10-bit choices are independent of colorspace.","default":"auto","multiselect":false,"options":["auto",8,10]}],"color_space":["COMBO",{"tooltip":"Colorspace of the input images. HDR selects BT.2020/HLG and HDR PQ selects BT.2020/PQ.","default":"sRGB","multiselect":false,"options":["sRGB","HDR","HDR PQ"]}],"codec":["COMBO",{"tooltip":"Optionally encode the video immediately. None keeps the images in tensor form; Auto uses H.264.","advanced":true,"default":"none","multiselect":false,"options":["none","auto","h264","av1"]}]}},"input_order":{"required":["images","fps"],"optional":["audio","bit_depth","color_space","codec"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["VIDEO"],"output_tooltips":[null],"output_matchtypes":null,"name":"CreateVideo","display_name":"Create Video","description":"Create a video from images.","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["images to video"],"essentials_category":"Video Tools","has_intermediate_output":false},"CrossViewPrepareClip":{"input":{"required":{"video":["VIDEO"],"duration":["FLOAT",{"default":5.0,"min":0.2,"max":15.0,"step":0.1,"tooltip":"Seconds of the clip to use, from its start. Snapped to the nearest frame count H3 accepts (17k+5 at 24 fps) and capped at the clip."}],"aspect_ratio":[["source","16:9","9:16","1:1","4:3","3:4","21:9"],{"default":"source","tooltip":"Output shape. 'source' keeps the clip's own ratio; the others centre-crop to it."}],"megapixels":["FLOAT",{"default":0.5,"min":0.1,"max":4.0,"step":0.05}],"multiple":["INT",{"default":32,"min":8,"max":128,"step":8}]}},"input_order":{"required":["video","duration","aspect_ratio","megapixels","multiple"]},"is_input_list":false,"output":["IMAGE","INT","INT","INT","AUDIO"],"output_is_list":[false,false,false,false,false],"output_name":["frames","width","height","length","audio"],"name":"CrossViewPrepareClip","display_name":"CrossView Prepare Clip","description":"","python_module":"custom_nodes.ComfyUI-CrossViewWarp","category":"CrossView","output_node":false,"has_intermediate_output":false,"output_tooltips":["The frames both stages use, at 24 fps.","Frame width.","Frame height.","Frame count, on H3's 17k+5 grid.","The clip's own audio over exactly those frames, silence if it has none."],"search_aliases":[]},"CrossViewWarp":{"input":{"required":{"frames":["IMAGE",{"tooltip":"Input video frames. Feed the same frames you send to the depth node."}],"azimuth":["FLOAT",{"default":-30.0,"min":-180.0,"max":180.0,"step":1.0,"tooltip":"Horizontal orbit angle (deg). Negative orbits LEFT, positive RIGHT. The strongest control: +-45 is where the model has been checked, and the training set covers out to +-90 evenly. Past 90 the hidden side is invented."}],"elevation":["FLOAT",{"default":20.0,"min":-90.0,"max":90.0,"step":1.0,"tooltip":"Vertical orbit angle (deg). Positive rises above the subject, negative looks up from below. UP is the weak direction: only 3.1% of the training set sits below -20."}],"distance":["FLOAT",{"default":1.0,"min":0.1,"max":3.0,"step":0.05,"tooltip":"Camera distance (1.0 = same as the source). Below 1 moves closer, above 1 pulls back - and pulling back reveals area the source never framed, so the magenta holes grow. The current LoRA does not follow this reliably."}],"hfov":["FLOAT",{"default":50.0,"min":0.0,"max":120.0,"step":1.0,"tooltip":"Assumed horizontal field of view (deg). ~50 is a normal lens; change it if the clip is clearly wide-angle or telephoto. 0 takes it from moge_geometry, which runs about 10% short - set it yourself if you know the lens."}],"vertical_shift":["FLOAT",{"default":0.0,"min":-1.0,"max":1.0,"step":0.02,"tooltip":"Vertical lens shift, as a fraction of image height. Positive moves the picture DOWN, i.e. the framing up, which rescues a clipped head. The camera does not move, so parallax is unchanged. Leave at 0 unless the framing needs it."}],"depth_ratio":["FLOAT",{"default":6.0,"min":1.5,"max":1000.0,"step":0.5,"tooltip":"Max far/near depth ratio of the scene. Lower = flatter relief and a cleaner warp; higher = more parallax but messier on cluttered scenes. Close-up faces: 2.5-4; mid shots: 4-8; deep/wide scenes: 8-16."}],"smooth_depth":["BOOLEAN",{"default":false,"tooltip":"Edge-aware depth smoothing before warping. Turn ON if the warp has too many speckle holes; it trades a little sharpness for cleaner disocclusion."}],"invert_depth":["BOOLEAN",{"default":false,"tooltip":"Flip depth polarity (near<->far). Leave OFF for the standard DA-V2 node. Turn ON only if the warp looks inside-out - background moving like foreground, or the subject caving in."}]},"optional":{"roll_lock":["BOOLEAN",{"default":true,"tooltip":"Retired and ignored. It rolled the camera to hold the subject's projected lean, but the orbit introduces no roll to correct - a level source gives a level target at every angle - and what it was reacting to is keystone from elevation, which a roll cannot fix. Both training datasets have exactly 0 camera roll."}],"pivot_override":["BOOLEAN",{"default":true,"tooltip":"Orbit around the explicit pivot below instead of an auto- estimated one. ON by default, which is the recommended setup."}],"pivot_x":["FLOAT",{"default":0.0,"min":-1000.0,"max":1000.0,"step":0.01,"tooltip":"Pivot X in the source camera frame (+ = right), in depth units. 0 = image centre. The camera aims at the pivot, so moving this also reframes the shot - unless keep_source_aim is ON, where it only moves the orbit centre."}],"pivot_y":["FLOAT",{"default":0.0,"min":-1000.0,"max":1000.0,"step":0.01,"tooltip":"Pivot Y (+ = down), in depth units. 0 = image centre. Doubles as a reframe like pivot_x. Use vertical_shift to reframe without moving the camera."}],"pivot_z":["FLOAT",{"default":1.05,"min":0.01,"max":1000.0,"step":0.01,"tooltip":"Pivot depth. ~1.05 sits on the nearest subject in the middle of the frame; raise it to orbit something further back. The orbit radius is distance * |pivot|, so this widens the arc as well as moving its centre."}],"use_keyframes":["BOOLEAN",{"default":false,"tooltip":"Animate the camera along the keyframes path instead of holding one pose. Switched on for you when the first keyframe is placed."}],"frame_count":["INT",{"default":0,"min":0,"max":100000,"step":1,"tooltip":"Retired and ignored. The clip length comes from the cached preview."}],"keyframes":["STRING",{"default":"","multiline":false,"tooltip":"Camera path as JSON - normally written for you by the KEY button on the preview, or by right-clicking the orbit sphere. Each entry: 'f' is the frame number counted from 1, 'az'/'el' are degrees, 'dist' is the distance. Optional 'vs' (vertical shift) and 'px'/'py'/'pz' (the pivot) inherit the static widgets when left out; px/py/pz need pivot_override ON. Example: [{\"f\":1,\"az\":-30,\"el\":20,\"dist\":1.0},{\"f\":49,\"az\":45,\"el\":10,\"dist\":1.2}] . The pose is held before the first and after the last keyframe, so a path may cover only part of the clip. Empty = no move."}],"interp_motion":[["linear","ease_in","ease_out","ease_in_out","smooth"],{"default":"linear","tooltip":"Shape and timing of the camera path. linear = straight legs at constant speed. ease_in / ease_out / ease_in_out = the same legs with the camera settling into every keyframe. smooth = a Catmull-Rom spline gliding through them with no corner. Only used while use_keyframes is ON."}],"interpolation":[["linear","smooth"],{"default":"linear","tooltip":"Driven by interp_motion."}],"keep_source_aim":["BOOLEAN",{"default":false,"tooltip":"Which point the orbiting camera looks at. OFF: it looks at the pivot, so the pivot sits dead centre and moving it reframes the shot. ON: it looks where the SOURCE camera looked, so your original framing is kept and the pivot is only the orbit centre. Needs pivot_x or pivot_y off zero to do anything - on the optical axis both aims are the same point. ON is what the training data does and scores 28.9 against 42.3 measured on the training warps; OFF by default for compatibility, turn it ON for new work."}],"depth":["IMAGE",{"tooltip":"Depth map for the same frames, from a Depth Anything V2 node. Brightness = depth; polarity is handled by invert_depth. Leave unconnected if you are feeding moge_geometry instead."}],"moge_geometry":["MOGE_GEOMETRY",{"tooltip":"Metric geometry from Run MoGe Inference, used instead of the depth image. Depth then arrives in real metres, so depth_ratio, invert_depth and smooth_depth no longer apply, and pivot and distance become real distances. Measured against ground truth it roughly halves the warp error. hfov comes from its intrinsics when hfov is 0."}],"camera_info":["LOAD3D_CAMERA",{"tooltip":"Exact target camera, from Create Camera Info or a Load 3D viewport. When connected the pose comes from it, and the orbit controls - angles, distance, hfov, the pivot and the keyframe path - are all ignored."}],"preview_size":["INT",{"default":512,"min":128,"max":768,"step":32,"tooltip":"Longest side the clip is cached at for the live preview. The warp is scale-invariant, so this trades sharpness and memory for frame rate: about 1 MB per frame and 18 fps at 384, 1.8 MB and 11 fps at 512. Takes effect on the next run."}],"frame_index":["INT",{"default":1,"min":1,"max":100000,"step":1,"tooltip":"Playhead for the live preview, counted from 1. The scrub bar under the preview drives it; it does not affect the output and needs no re-run."}]},"hidden":{"unique_id":"UNIQUE_ID"}},"input_order":{"required":["frames","azimuth","elevation","distance","hfov","vertical_shift","depth_ratio","smooth_depth","invert_depth"],"optional":["roll_lock","pivot_override","pivot_x","pivot_y","pivot_z","use_keyframes","frame_count","keyframes","interp_motion","interpolation","keep_source_aim","depth","moge_geometry","camera_info","preview_size","frame_index"],"hidden":["unique_id"]},"is_input_list":false,"output":["IMAGE","IMAGE"],"output_is_list":[false,false],"output_name":["warp","orbit_view"],"name":"CrossViewWarp","display_name":"CrossView Warp","description":"","python_module":"custom_nodes.ComfyUI-CrossViewWarp","category":"CrossView","output_node":true,"has_intermediate_output":false,"output_tooltips":["The warp control video (magenta = disoccluded holes) -> connect to the IC-LoRA reference guide for the WARP stream.","Front view of the orbit globe with the camera marker at the requested az/el/dist - wire to a PreviewImage/SaveImage to document the camera setup."],"search_aliases":[]},"CustomCombo":{"input":{"required":{"choice":["COMBO",{"multiselect":false,"options":[]}]}},"input_order":{"required":["choice"]},"is_input_list":false,"output":["STRING","INT"],"output_is_list":[false,false],"output_name":["STRING","INDEX"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"CustomCombo","display_name":"Custom Combo","description":"","python_module":"comfy_extras.nodes_logic","category":"utilities","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptyAudio":{"input":{"required":{"duration":["FLOAT",{"tooltip":"Duration of the empty audio clip in seconds","default":60.0,"min":0.0,"max":18446744073709551615,"step":0.01}],"sample_rate":["INT",{"tooltip":"Sample rate of the empty audio clip.","advanced":true,"default":44100,"min":1,"max":192000}],"channels":["INT",{"tooltip":"Number of audio channels (1 for mono, 2 for stereo).","advanced":true,"default":2,"min":1,"max":2}]}},"input_order":{"required":["duration","sample_rate","channels"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyAudio","display_name":"Empty Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["blank audio"],"essentials_category":null,"has_intermediate_output":false},"EmptyFlux2LatentImage":{"input":{"required":{"width":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["width","height","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyFlux2LatentImage","display_name":"Empty Flux 2 Latent","description":"","python_module":"comfy_extras.nodes_flux","category":"model/latent/flux","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptyImage":{"input":{"required":{"width":["INT",{"default":512,"min":1,"max":16384,"step":1}],"height":["INT",{"default":512,"min":1,"max":16384,"step":1}],"batch_size":["INT",{"default":1,"min":1,"max":4096}],"color":["INT",{"default":0,"min":0,"max":16777215,"step":1,"display":"color"}]}},"input_order":{"required":["width","height","batch_size","color"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"EmptyImage","display_name":"Empty Image","description":"","python_module":"nodes","category":"image","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"EmptyLTXVLatentVideo":{"input":{"required":{"width":["INT",{"default":768,"min":64,"max":16384,"step":32}],"height":["INT",{"default":512,"min":64,"max":16384,"step":32}],"length":["INT",{"default":97,"min":1,"max":16384,"step":8}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["width","height","length","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyLTXVLatentVideo","display_name":"Empty LTXV Latent Video","description":"","python_module":"comfy_extras.nodes_lt","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptyLatentAudio":{"input":{"required":{"seconds":["FLOAT",{"default":47.6,"min":1.0,"max":1000.0,"step":0.1}],"batch_size":["INT",{"tooltip":"The number of latent images in the batch.","default":1,"min":1,"max":4096}]}},"input_order":{"required":["seconds","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyLatentAudio","display_name":"Empty Latent Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":"Audio","has_intermediate_output":false},"EmptyLatentImage":{"input":{"required":{"width":["INT",{"default":1024,"min":16,"max":16384,"step":8,"tooltip":"The width of the latent images in pixels."}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":8,"tooltip":"The height of the latent images in pixels."}],"batch_size":["INT",{"default":1,"min":1,"max":4096,"tooltip":"The number of latent images in the batch."}]}},"input_order":{"required":["width","height","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"EmptyLatentImage","display_name":"Empty Latent Image","description":"Create a new batch of empty latent images to be denoised via sampling.","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"output_tooltips":["The empty latent image batch."],"search_aliases":["empty","empty latent","new latent","create latent","blank latent","blank"]},"EmptyMiniMaxMusic3LatentAudio":{"input":{"required":{"seconds":["FLOAT",{"default":120.0,"min":0.04,"max":360.0,"step":0.04}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["seconds","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptyMiniMaxMusic3LatentAudio","display_name":"Empty MiniMax Music3 Latent Audio","description":"Creates an empty MiniMax Music3 audio latent for the requested duration.","python_module":"comfy_extras.nodes_minimax_music","category":"model/latent/minimax music","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"EmptySD3LatentImage":{"input":{"required":{"width":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":16}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["width","height","batch_size"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"EmptySD3LatentImage","display_name":null,"description":"","python_module":"comfy_extras.nodes_sd3","category":"model/latent/stable diffusion","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"Flux3LayoutPrompt":{"input":{"required":{"prompt":["STRING",{"tooltip":"The prompt text. Cite box N as <obj_N> to tie words to it.","multiline":true,"dynamicPrompts":true}],"element":["ARRAY",{"tooltip":"Prompt elements from the node Create Bounding Boxes. Colour palettes are not used, describe colours in the description."}]}},"input_order":{"required":["prompt","element"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["prompt"],"output_tooltips":[null],"output_matchtypes":null,"name":"Flux3LayoutPrompt","display_name":"FLUX 3 Image Layout Prompt","description":"Appends the boxes from Create Bounding Boxes to a FLUX 3 Image prompt as its layout table.","python_module":"comfy_extras.nodes_flux","category":"model/conditioning/flux","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"FluxKontextImageScale":{"input":{"required":{"image":["IMAGE",{}]}},"input_order":{"required":["image"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"FluxKontextImageScale","display_name":null,"description":"This node resizes the image to one that is more optimal for flux kontext.","python_module":"comfy_extras.nodes_flux","category":"model/conditioning/flux","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"FluxKontextMultiReferenceLatentMethod":{"input":{"required":{"conditioning":["CONDITIONING",{}],"reference_latents_method":["COMBO",{"advanced":true,"multiselect":false,"options":["offset","index","uxo/uno","index_timestep_zero"]}]}},"input_order":{"required":["conditioning","reference_latents_method"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"output_tooltips":[null],"output_matchtypes":null,"name":"FluxKontextMultiReferenceLatentMethod","display_name":"Edit Model Reference Method","description":"","python_module":"comfy_extras.nodes_flux","category":"model/conditioning/flux","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"FrameInterpolate":{"input":{"required":{"interp_model":["INTERP_MODEL",{}],"images":["IMAGE",{}],"multiplier":["INT",{"default":2,"min":2,"max":16}]}},"input_order":{"required":["interp_model","images","multiplier"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"FrameInterpolate","display_name":"Run Frame Interpolation Model","description":"","python_module":"comfy_extras.nodes_frame_interpolation","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["rife","film","frame interpolation","slow motion","interpolate frames","vfi"],"essentials_category":null,"has_intermediate_output":false},"FrameInterpolationModelLoader":{"input":{"required":{"model_name":["COMBO",{"tooltip":"Select a frame interpolation model to load. Models must be placed in the 'frame_interpolation' folder.","multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["model_name"]},"is_input_list":false,"output":["INTERP_MODEL"],"output_is_list":[false],"output_name":["INTERP_MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"FrameInterpolationModelLoader","display_name":"Load Frame Interpolation Model","description":"","python_module":"comfy_extras.nodes_frame_interpolation","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"GemmaAPITextEncode":{"input":{"required":{"api_key":["STRING",{"default":"","placeholder":"API_KEY","multiline":false,"tooltip":"API key for authentication"}],"prompt":["STRING",{"multiline":true,"default":"","tooltip":"Text prompt to encode"}],"enhance_prompt":["BOOLEAN",{"default":true,"tooltip":"When enabled, the prompt is enhanced using Gemma 3 before encoding"}],"ckpt_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"],{"tooltip":"The model to read the API text-encoder selector from. Either a checkpoint or a diffusion model."}]}},"input_order":{"required":["api_key","prompt","enhance_prompt","ckpt_name"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["conditioning"],"name":"GemmaAPITextEncode","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Gemma API Text Encode","description":"","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"api node/text/Lightricks","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"GetImageSize":{"input":{"required":{"image":["IMAGE",{}]},"hidden":{"unique_id":["UNIQUE_ID"]}},"input_order":{"required":["image"],"hidden":["unique_id"]},"is_input_list":false,"output":["INT","INT","INT"],"output_is_list":[false,false,false],"output_name":["width","height","batch_size"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"GetImageSize","display_name":"Get Image Size","description":"Returns the width, height, and batch size of the image.","python_module":"comfy_extras.nodes_images","category":"image","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["dimensions","resolution","image info"],"essentials_category":null,"has_intermediate_output":false},"GetVideoComponents":{"input":{"required":{"video":["VIDEO",{"tooltip":"The video to extract components from."}]}},"input_order":{"required":["video"]},"is_input_list":false,"output":["IMAGE","AUDIO","FLOAT","COMBO","COMBO"],"output_is_list":[false,false,false,false,false],"output_name":["images","audio","fps","bit_depth","color_space"],"output_tooltips":[null,null,null,null,null],"output_matchtypes":null,"name":"GetVideoComponents","display_name":"Get Video Components","description":"Extracts video frames, audio, frame rate, bit depth, and color space.","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["extract frames","split video","video to images","demux"],"essentials_category":null,"has_intermediate_output":false},"ImageBatch":{"input":{"required":{"image1":["IMAGE"],"image2":["IMAGE"]}},"input_order":{"required":["image1","image2"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"ImageBatch","display_name":"Batch Images (DEPRECATED)","description":"","python_module":"nodes","category":"image/batch","output_node":false,"has_intermediate_output":false,"deprecated":true,"search_aliases":["combine images","merge images","stack images"]},"ImageColorSpace":{"input":{"required":{"image":["IMAGE",{}],"source":["COMBO",{"tooltip":"Color space of the input pixels.","default":"sRGB","multiselect":false,"options":["sRGB","HDR","HDR PQ","linear","HDR LogC3","HDR ACEScct"]}],"destination":["COMBO",{"tooltip":"Color space of the output pixels. Set the save node to this same color space. Convert LogC3 or ACEScct to linear before saving EXR, or to sRGB/HDR/HDR PQ before saving video.","default":"sRGB","multiselect":false,"options":["sRGB","HDR","HDR PQ","linear","HDR LogC3","HDR ACEScct"]}]}},"input_order":{"required":["image","source","destination"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageColorSpace","display_name":"Convert Image Color Space","description":"Convert sRGB, linear Rec.709, HDR (Rec.2020 HLG), HDR PQ (Rec.2020 PQ), HDR LogC3, and HDR ACEScct. LogC3 uses the EI 800 curve with Rec.709 primaries and codes clamped to [0, 1]. ACEScct uses AP1 primaries and D60 white, with Bradford adaptation to D65. Convert LogC3 or ACEScct to linear for EXR saving. Linear 1.0 uses the same 203-nit reference white as sRGB; HLG uses a 1000-nit reference display. Linear and ACEScct outputs preserve extended values. Linear-to-HDR conversions preserve highlights without tone mapping; HLG and PQ outputs clip negative channels. SDR output and PQ-to-HLG conversion tone-map excess luminance across the batch and compress out-of-gamut colors. Conversions compute in float32 and return the intermediate device and dtype. Straight alpha is not color-transformed.","python_module":"comfy_extras.nodes_images","category":"image/color","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"ImageCompare":{"input":{"required":{"compare_view":["IMAGECOMPARE",{"socketless":true}]},"optional":{"image_a":["IMAGE",{}],"image_b":["IMAGE",{}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["compare_view"],"optional":["image_a","image_b"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":[],"output_is_list":[],"output_name":[],"output_tooltips":[],"output_matchtypes":null,"name":"ImageCompare","display_name":"Compare Images","description":"Compares two images side by side with a slider.","python_module":"comfy_extras.nodes_image_compare","category":"image","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":"Image Tools","has_intermediate_output":false},"ImageCompositeMasked":{"input":{"required":{"destination":["IMAGE",{}],"source":["IMAGE",{}],"x":["INT",{"default":0,"min":0,"max":16384,"step":1}],"y":["INT",{"default":0,"min":0,"max":16384,"step":1}],"resize_source":["BOOLEAN",{"default":false}]},"optional":{"mask":["MASK",{}]}},"input_order":{"required":["destination","source","x","y","resize_source"],"optional":["mask"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageCompositeMasked","display_name":"Image Composite Masked","description":"","python_module":"comfy_extras.nodes_mask","category":"image/compositing","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["overlay","layer","paste image","images composition"],"essentials_category":null,"has_intermediate_output":false},"ImageFromBatch":{"input":{"required":{"image":["IMAGE",{}],"batch_index":["INT",{"default":0,"min":-16384,"max":16384}],"length":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["image","batch_index","length"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageFromBatch","display_name":"Get Image from Batch","description":"","python_module":"comfy_extras.nodes_images","category":"image/batch","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["select image","pick from batch","extract image"],"essentials_category":null,"has_intermediate_output":false},"ImageInvert":{"input":{"required":{"image":["IMAGE"]}},"input_order":{"required":["image"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"ImageInvert","display_name":"Invert Image Colors","description":"","python_module":"nodes","category":"image/color","output_node":false,"has_intermediate_output":false,"search_aliases":["reverse colors"],"essentials_category":"Image Tools"},"ImagePadForOutpaint":{"input":{"required":{"image":["IMAGE"],"left":["INT",{"default":0,"min":0,"max":16384,"step":8}],"top":["INT",{"default":0,"min":0,"max":16384,"step":8}],"right":["INT",{"default":0,"min":0,"max":16384,"step":8}],"bottom":["INT",{"default":0,"min":0,"max":16384,"step":8}],"feathering":["INT",{"default":40,"min":0,"max":16384,"step":1,"advanced":true}]}},"input_order":{"required":["image","left","top","right","bottom","feathering"]},"is_input_list":false,"output":["IMAGE","MASK"],"output_is_list":[false,false],"output_name":["IMAGE","MASK"],"name":"ImagePadForOutpaint","display_name":"Pad Image for Outpainting","description":"","python_module":"nodes","category":"image/transform","output_node":false,"has_intermediate_output":false,"search_aliases":["extend canvas","expand image"]},"ImageScale":{"input":{"required":{"image":["IMAGE"],"upscale_method":[["nearest-exact","bilinear","area","bicubic","lanczos"]],"width":["INT",{"default":512,"min":0,"max":16384,"step":1}],"height":["INT",{"default":512,"min":0,"max":16384,"step":1}],"crop":[["disabled","center"]]}},"input_order":{"required":["image","upscale_method","width","height","crop"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"ImageScale","display_name":"Upscale Image","description":"","python_module":"nodes","category":"image/upscaling","output_node":false,"has_intermediate_output":false,"search_aliases":["resize","resize image","scale image","image resize","zoom","zoom in","change size"],"essentials_category":"Image Tools"},"ImageScaleToTotalPixels":{"input":{"required":{"image":["IMAGE",{}],"upscale_method":["COMBO",{"multiselect":false,"options":["nearest-exact","bilinear","area","bicubic","lanczos"]}],"megapixels":["FLOAT",{"default":1.0,"min":0.01,"max":16.0,"step":0.01}],"resolution_steps":["INT",{"advanced":true,"default":1,"min":1,"max":256}]}},"input_order":{"required":["image","upscale_method","megapixels","resolution_steps"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageScaleToTotalPixels","display_name":"Scale Image to Total Pixels","description":"","python_module":"comfy_extras.nodes_post_processing","category":"image/upscaling","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"ImageStitch":{"input":{"required":{"image1":["IMAGE",{}],"direction":["COMBO",{"default":"right","multiselect":false,"options":["right","down","left","up"]}],"match_image_size":["BOOLEAN",{"default":true}],"spacing_width":["INT",{"advanced":true,"default":0,"min":0,"max":1024,"step":2}],"spacing_color":["COMBO",{"advanced":true,"default":"white","multiselect":false,"options":["white","black","red","green","blue"]}]},"optional":{"image2":["IMAGE",{}]}},"input_order":{"required":["image1","direction","match_image_size","spacing_width","spacing_color"],"optional":["image2"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageStitch","display_name":"Stitch Images","description":"Stitches image2 to image1 in the specified direction.\nIf image2 is not provided, returns image1 unchanged.\nOptional spacing can be added between images.","python_module":"comfy_extras.nodes_images","category":"image/transform","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["combine images","join images","concatenate images","side by side"],"essentials_category":null,"has_intermediate_output":false},"ImageToMask":{"input":{"required":{"image":["IMAGE",{}],"channel":["COMBO",{"multiselect":false,"options":["red","green","blue","alpha"]}]}},"input_order":{"required":["image","channel"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["MASK"],"output_tooltips":[null],"output_matchtypes":null,"name":"ImageToMask","display_name":"Convert Image to Mask","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["extract channel","channel to mask"],"essentials_category":null,"has_intermediate_output":false},"InvertMask":{"input":{"required":{"mask":["MASK",{}]}},"input_order":{"required":["mask"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["MASK"],"output_tooltips":[null],"output_matchtypes":null,"name":"InvertMask","display_name":"Invert Mask","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["reverse mask","flip mask"],"essentials_category":null,"has_intermediate_output":false},"JoinImageWithAlpha":{"input":{"required":{"image":["IMAGE",{}],"alpha":["MASK",{}]}},"input_order":{"required":["image","alpha"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"JoinImageWithAlpha","display_name":"Join Image with Alpha","description":"","python_module":"comfy_extras.nodes_compositing","category":"image/compositing","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["add transparency","apply alpha","composite alpha","RGBA"],"essentials_category":null,"has_intermediate_output":false},"JsonExtractString":{"input":{"required":{"json_string":["STRING",{"multiline":true}],"key":["STRING",{"multiline":false}]}},"input_order":{"required":["json_string","key"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"JsonExtractString","display_name":"Extract Text from JSON","description":"Extract a key's value from the first valid JSON object in the text, including Markdown code blocks.","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["json","extract json","parse json","json value","read json"],"essentials_category":null,"has_intermediate_output":false},"KSampler":{"input":{"required":{"model":["MODEL",{"tooltip":"The model used for denoising the input latent."}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true,"tooltip":"The random seed used for creating the noise."}],"steps":["INT",{"default":20,"min":1,"max":10000,"tooltip":"The number of steps used in the denoising process."}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01,"tooltip":"The Classifier-Free Guidance scale balances creativity and adherence to the prompt. Higher values result in images more closely matching the prompt however too high values will negatively impact quality."}],"sampler_name":[["euler","euler_cfg_pp","euler_ancestral","euler_ancestral_cfg_pp","heun","heunpp2","exp_heun_2_x0","exp_heun_2_x0_sde","dpm_2","dpm_2_ancestral","lms","dpm_fast","dpm_adaptive","dpmpp_2s_ancestral","dpmpp_2s_ancestral_cfg_pp","dpmpp_sde","dpmpp_sde_gpu","dpmpp_2m","dpmpp_2m_cfg_pp","dpmpp_2m_sde","dpmpp_2m_sde_gpu","dpmpp_2m_sde_heun","dpmpp_2m_sde_heun_gpu","dpmpp_3m_sde","dpmpp_3m_sde_gpu","ddpm","lcm","ipndm","ipndm_v","deis","cfgpp_ud10_ab","res_multistep","res_multistep_cfg_pp","res_multistep_ancestral","res_multistep_ancestral_cfg_pp","gradient_estimation","gradient_estimation_cfg_pp","er_sde","seeds_2","seeds_3","sa_solver","sa_solver_pece","ddim","uni_pc","uni_pc_bh2"],{"tooltip":"The algorithm used when sampling, this can affect the quality, speed, and style of the generated output."}],"scheduler":[["simple","sgm_uniform","karras","exponential","ddim_uniform","beta","normal","linear_quadratic","kl_optimal"],{"tooltip":"The scheduler controls how noise is gradually removed to form the image."}],"positive":["CONDITIONING",{"tooltip":"The conditioning describing the attributes you want to include in the image."}],"negative":["CONDITIONING",{"tooltip":"The conditioning describing the attributes you want to exclude from the image."}],"latent_image":["LATENT",{"tooltip":"The latent image to denoise."}],"denoise":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01,"tooltip":"The amount of denoising applied, lower values will maintain the structure of the initial image allowing for image to image sampling."}]}},"input_order":{"required":["model","seed","steps","cfg","sampler_name","scheduler","positive","negative","latent_image","denoise"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"KSampler","display_name":"KSampler","description":"Uses the provided model, positive and negative conditioning to denoise the latent image.","python_module":"nodes","category":"model/sampling","output_node":false,"has_intermediate_output":false,"output_tooltips":["The denoised latent."],"search_aliases":["sampler","sample","generate","denoise","diffuse","txt2img","img2img"]},"KSamplerAdvanced":{"input":{"required":{"model":["MODEL"],"add_noise":[["enable","disable"],{"advanced":true}],"noise_seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}],"steps":["INT",{"default":20,"min":1,"max":10000}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"sampler_name":[["euler","euler_cfg_pp","euler_ancestral","euler_ancestral_cfg_pp","heun","heunpp2","exp_heun_2_x0","exp_heun_2_x0_sde","dpm_2","dpm_2_ancestral","lms","dpm_fast","dpm_adaptive","dpmpp_2s_ancestral","dpmpp_2s_ancestral_cfg_pp","dpmpp_sde","dpmpp_sde_gpu","dpmpp_2m","dpmpp_2m_cfg_pp","dpmpp_2m_sde","dpmpp_2m_sde_gpu","dpmpp_2m_sde_heun","dpmpp_2m_sde_heun_gpu","dpmpp_3m_sde","dpmpp_3m_sde_gpu","ddpm","lcm","ipndm","ipndm_v","deis","cfgpp_ud10_ab","res_multistep","res_multistep_cfg_pp","res_multistep_ancestral","res_multistep_ancestral_cfg_pp","gradient_estimation","gradient_estimation_cfg_pp","er_sde","seeds_2","seeds_3","sa_solver","sa_solver_pece","ddim","uni_pc","uni_pc_bh2"]],"scheduler":[["simple","sgm_uniform","karras","exponential","ddim_uniform","beta","normal","linear_quadratic","kl_optimal"]],"positive":["CONDITIONING"],"negative":["CONDITIONING"],"latent_image":["LATENT"],"start_at_step":["INT",{"default":0,"min":0,"max":10000,"advanced":true}],"end_at_step":["INT",{"default":10000,"min":0,"max":10000,"advanced":true}],"return_with_leftover_noise":[["disable","enable"],{"advanced":true}]}},"input_order":{"required":["model","add_noise","noise_seed","steps","cfg","sampler_name","scheduler","positive","negative","latent_image","start_at_step","end_at_step","return_with_leftover_noise"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"KSamplerAdvanced","display_name":"KSampler (Advanced)","description":"","python_module":"nodes","category":"model/sampling","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"KSamplerSelect":{"input":{"required":{"sampler_name":["COMBO",{"multiselect":false,"options":["euler","euler_cfg_pp","euler_ancestral","euler_ancestral_cfg_pp","heun","heunpp2","exp_heun_2_x0","exp_heun_2_x0_sde","dpm_2","dpm_2_ancestral","lms","dpm_fast","dpm_adaptive","dpmpp_2s_ancestral","dpmpp_2s_ancestral_cfg_pp","dpmpp_sde","dpmpp_sde_gpu","dpmpp_2m","dpmpp_2m_cfg_pp","dpmpp_2m_sde","dpmpp_2m_sde_gpu","dpmpp_2m_sde_heun","dpmpp_2m_sde_heun_gpu","dpmpp_3m_sde","dpmpp_3m_sde_gpu","ddpm","lcm","ipndm","ipndm_v","deis","cfgpp_ud10_ab","res_multistep","res_multistep_cfg_pp","res_multistep_ancestral","res_multistep_ancestral_cfg_pp","gradient_estimation","gradient_estimation_cfg_pp","er_sde","seeds_2","seeds_3","sa_solver","sa_solver_pece","ddim","uni_pc","uni_pc_bh2"]}]}},"input_order":{"required":["sampler_name"]},"is_input_list":false,"output":["SAMPLER"],"output_is_list":[false],"output_name":["SAMPLER"],"output_tooltips":[null],"output_matchtypes":null,"name":"KSamplerSelect","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/samplers","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXAddVideoICLoRAGuide":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"latent":["LATENT",{"tooltip":"Video-only latent to condition. Must be a 5D video latent (batch, channels, frames, height, width)."}],"image":["IMAGE",{}],"frame_idx":["INT",{"tooltip":"RoPE position of the reference, in pixel frames.\n0 aligns the reference with the generated clip. Use for video references.\n>0 places the reference at a specific pixel-frame position in the clip.\n-1 places the reference outside the generated frame range, while still using it for attention conditioning. Use for single-image references.\nFor multi-frame references, >0 values are rounded down to 1 mod 8.","default":0,"min":-9999,"max":9999}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}],"latent_downscale_factor":["FLOAT",{"tooltip":"For IC-LoRA on small grid. 1 means original size, 2 means half size, 3 means third, etc.","default":1.0,"min":1.0,"max":10.0,"step":1.0}],"crop":["COMBO",{"tooltip":"Crop mode when resizing. 'center' crops to fit, 'disabled' stretches to fit.","default":"disabled","multiselect":false,"options":["disabled","center"]}],"use_tiled_encode":["BOOLEAN",{"tooltip":"Enable tiled VAE encoding for large resolutions/long videos to reduce memory usage.","default":false}],"tile_size":["INT",{"tooltip":"Spatial tile size for tiled encoding. Only used when use_tiled_encode is enabled.","default":256,"min":64,"max":512,"step":32}],"tile_overlap":["INT",{"tooltip":"Overlap between tiles for tiled encoding. Only used when use_tiled_encode is enabled.","default":64,"min":16,"max":256,"step":16}]},"optional":{"use_streaming":["BOOLEAN",{"tooltip":"Off (default): encode the whole clip as one IC-LoRA guide. On: re-encode one fresh guide per LTXVTiledFusionSampler temporal window and write the plan the sampler reads. Set tile_frames on this node and the sampler to the same value (97 for LTX).","default":false}],"tile_frames":["INT",{"tooltip":"Window length in pixel frames, used only when use_streaming is on. Snapped to 8n+1. Must match LTXVTiledFusionSampler. Ignored when use_streaming is off.","default":97,"min":0,"max":1000,"step":1}]}},"input_order":{"required":["positive","negative","vae","latent","image","frame_idx","strength","latent_downscale_factor","crop","use_tiled_encode","tile_size","tile_overlap"],"optional":["use_streaming","tile_frames"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXAddVideoICLoRAGuide","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Add Video IC-LoRA Guide","description":"Adds one or more conditioning frames starting at the specified frame index. Supports both single images and multi-frame videos. The latent_downscale_factor resizes input to a fraction of the target size (1 = original, 2 = half, 3 = third, etc.) for IC-LoRA on small grids. For LTXVTiledFusionSampler temporal windows, turn on use_streaming and set tile_frames to the same 8n+1 value as the sampler (97 for LTX).","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXAddVideoICLoRAGuideAdvanced":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"latent":["LATENT",{"tooltip":"Video-only latent to condition. Must be a 5D video latent."}],"image":["IMAGE",{}],"frame_idx":["INT",{"tooltip":"RoPE position of the reference, in pixel frames.\n0 aligns the reference with the generated clip. Use for video references.\n>0 places the reference at a specific pixel-frame position in the clip.\n-1 places the reference outside the generated frame range, while still using it for attention conditioning. Use for single-image references.\nFor multi-frame references, >0 values are rounded down to 1 mod 8.","default":0,"min":-9999,"max":9999}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}],"latent_downscale_factor":["FLOAT",{"tooltip":"For IC-LoRA on small grid. 1 = original size, 2 = half, etc.","default":1.0,"min":1.0,"max":10.0,"step":1.0}],"crop":["COMBO",{"default":"disabled","multiselect":false,"options":["disabled","center"]}],"use_tiled_encode":["BOOLEAN",{"default":false}],"tile_size":["INT",{"default":256,"min":64,"max":512,"step":32}],"tile_overlap":["INT",{"default":64,"min":16,"max":256,"step":16}],"attention_strength":["FLOAT",{"tooltip":"Controls how strongly this guide influences generation via self-attention. 1.0 = full conditioning (default), 0.0 = ignore. When an attention_mask is also provided, this multiplies the mask values.","default":1.0,"min":0.0,"max":1.0,"step":0.01}]},"optional":{"attention_mask":["MASK",{"tooltip":"Optional pixel-space spatial mask. Shape (F, H, W) or (H, W). Values in [0, 1]. Controls per-region conditioning influence. Multiplied by attention_strength."}],"use_streaming":["BOOLEAN",{"tooltip":"Off (default): whole-clip guide. On: one fresh encode per fusion temporal window. Must match the sampler's tile_frames.","default":false}],"tile_frames":["INT",{"tooltip":"Used only when use_streaming is on. Snapped to 8n+1. 97 for LTX. Must match the sampler.","default":97,"min":0,"max":1000,"step":1}]}},"input_order":{"required":["positive","negative","vae","latent","image","frame_idx","strength","latent_downscale_factor","crop","use_tiled_encode","tile_size","tile_overlap","attention_strength"],"optional":["attention_mask","use_streaming","tile_frames"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXAddVideoICLoRAGuideAdvanced","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Add Video IC-LoRA Guide Advanced","description":"Adds IC-LoRA guide conditioning with per-guide attention strength control. Same as LTXAddVideoICLoRAGuide, but allows controlling how strongly this guide influences generation via self-attention, optionally with a spatial mask. use_streaming + matching tile_frames re-encodes one guide per Tiled Fusion window.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXFloatToInt":{"input":{"required":{"a":["FLOAT",{"default":0.0}]}},"input_order":{"required":["a"]},"is_input_list":false,"output":["INT"],"output_is_list":[false],"output_name":["INT"],"name":"LTXFloatToInt","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Float To Int","description":"","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"math/conversion","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LTXICLoRALoaderModelOnly":{"input":{"required":{"model":["MODEL",{}],"lora_name":["COMBO",{"multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}],"strength_model":["FLOAT",{"default":1.0,"min":-100.0,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","lora_name","strength_model"]},"is_input_list":false,"output":["MODEL","FLOAT"],"output_is_list":[false,false],"output_name":["model","latent_downscale_factor"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"LTXICLoRALoaderModelOnly","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 IC-LoRA Loader Model Only","description":"Loads a LoRA model and extracts the latent_downscale_factor from the safetensors metadata.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVAddGuide":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"latent":["LATENT",{}],"image":["IMAGE",{"tooltip":"Image or video to condition the latent video on. Must be 8*n + 1 frames. If the video is not 8*n + 1 frames, it will be cropped to the nearest 8*n + 1 frames."}],"frame_idx":["INT",{"tooltip":"Frame index to start the conditioning at. For single-frame images or videos with 1-8 frames, any frame_idx value is acceptable. For videos with 9+ frames, frame_idx must be divisible by 8, otherwise it will be rounded down to the nearest multiple of 8. Negative values are counted from the end of the video.","default":0,"min":-9999,"max":9999}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":10.0,"step":0.01}]},"optional":{"attention_mask":["MASK",{"tooltip":"Optional pixel-space spatial mask. Controls per-region conditioning influence via self-attention, multiplied by strength."}],"iclora_parameters":["IC_LORA_PARAMETERS",{"tooltip":"Optional IC-LoRA parameters from a Get IC-LoRA Parameters node. Used for adjusting guide processing as required by certain IC-LoRAs (eg. those with a reference_downscale_factor > 1). When chained, each LTXVAddGuide uses only the parameters connected to it."}]}},"input_order":{"required":["positive","negative","vae","latent","image","frame_idx","strength"],"optional":["attention_mask","iclora_parameters"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXVAddGuide","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVAudioVAEDecode":{"input":{"required":{"samples":["LATENT",{"tooltip":"The latent to be decoded."}],"audio_vae":["VAE",{"display_name":"Audio VAE","tooltip":"The Audio VAE model used for decoding the latent."}]}},"input_order":{"required":["samples","audio_vae"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["Audio"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVAudioVAEDecode","display_name":"LTXV Audio VAE Decode","description":"","python_module":"comfy_extras.nodes_lt_audio","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVAudioVAEEncode":{"input":{"required":{"audio":["AUDIO",{"tooltip":"The audio to be encoded."}],"audio_vae":["VAE",{"display_name":"Audio VAE","tooltip":"The Audio VAE model to use for encoding."}]}},"input_order":{"required":["audio","audio_vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["Audio Latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVAudioVAEEncode","display_name":"LTXV Audio VAE Encode","description":"","python_module":"comfy_extras.nodes_lt_audio","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVConcatAVLatent":{"input":{"required":{"video_latent":["LATENT",{}],"audio_latent":["LATENT",{}]}},"input_order":{"required":["video_latent","audio_latent"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVConcatAVLatent","display_name":"Concat AV Latent","description":"Merge a video latent and an audio latent into a joint AV latent (any AV model, e.g. LTXV or MiniMax H3).","python_module":"comfy_extras.nodes_lt","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVConditioning":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"frame_rate":["FLOAT",{"default":25.0,"min":0.0,"max":1000.0,"step":0.01}]}},"input_order":{"required":["positive","negative","frame_rate"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING"],"output_is_list":[false,false],"output_name":["positive","negative"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"LTXVConditioning","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVCropGuides":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"latent":["LATENT",{}]}},"input_order":{"required":["positive","negative","latent"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"LTXVCropGuides","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVDualCFGGuider":{"input":{"required":{"model":["MODEL",{}],"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"video_cfg":["FLOAT",{"default":3.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"audio_cfg":["FLOAT",{"default":7.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}]}},"input_order":{"required":["model","positive","negative","video_cfg","audio_cfg"]},"is_input_list":false,"output":["GUIDER"],"output_is_list":[false],"output_name":["GUIDER"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVDualCFGGuider","display_name":"LTXV Dual CFG Guider","description":"Separate CFG scales for the video and audio modalities of a packed LTXV-AV latent.","python_module":"comfy_extras.nodes_lt","category":"model/sampling/guiders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVEmptyLatentAudio":{"input":{"required":{"frames_number":["INT",{"tooltip":"Number of frames.","default":97,"min":1,"max":1000,"step":1,"display":"number"}],"frame_rate":["FLOAT,INT",{"tooltip":"Number of frames per second.","default":25.0,"widgetType":"FLOAT","min":1.0,"max":1000.0,"step":0.01,"display":"number"}],"batch_size":["INT",{"tooltip":"The number of latent audio samples in the batch.","default":1,"min":1,"max":4096,"display":"number"}],"audio_vae":["VAE",{"display_name":"Audio VAE","tooltip":"The Audio VAE model to get configuration from."}]}},"input_order":{"required":["frames_number","frame_rate","batch_size","audio_vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["Latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVEmptyLatentAudio","display_name":"Empty LTXV Latent Audio","description":"","python_module":"comfy_extras.nodes_lt_audio","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVHDRDecodePostprocess":{"input":{"required":{"image":["IMAGE"]},"optional":{"transfer":[["acescct","logc3"],{"default":"acescct","tooltip":"Working-space curve of the VAE decode. Use acescct for SDR\u2192HDR IC-LoRA (after LTXVSDRToHDRWorkingSpace) and for native HDR EXR; logc3 only for legacy LogC3 HDR LoRA checkpoints."}],"exposure":["FLOAT",{"default":0.0,"min":-10.0,"max":10.0,"step":0.1,"display":"slider","tooltip":"Exposure in stops (EV). 0 = no change, +1 = 2x brighter, -1 = half brightness. Affects the tonemapped preview only."}],"save_exr":["BOOLEAN",{"default":false,"tooltip":"Save HDR frames as an EXR sequence."}],"exr_color_space":[["linear","acescct","srgb_linear","acescg"],{"default":"acescg","tooltip":"EXR contents when save_exr is on. Default 'acescg' is tagged ACEScg scene-linear. 'acescct' is ACEScct log; 'srgb_linear' is tagged Rec.709 linear; 'linear' is legacy raw decompress (untagged). All paths use OpenImageIO."}],"output_dir":["STRING",{"default":"output/hdr_exr","tooltip":"Directory for EXR frames (relative to ComfyUI output directory, or absolute path)."}],"filename_prefix":["STRING",{"default":"frame"}],"half_precision":["BOOLEAN",{"default":true,"tooltip":"Save EXR as float16 (half). Smaller files, negligible quality loss for most workflows."}]}},"input_order":{"required":["image"],"optional":["transfer","exposure","save_exr","exr_color_space","output_dir","filename_prefix","half_precision"]},"is_input_list":false,"output":["IMAGE","IMAGE"],"output_is_list":[false,false],"output_name":["tonemapped","hdr_linear"],"name":"LTXVHDRDecodePostprocess","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXVHDR Decode Postprocess","description":"Decompresses VAE-decoded HDR (ACEScct or legacy LogC3), tonemaps for preview, and optionally writes EXR (linear / ACEScct / sRGB linear / ACEScg). Wire hdr_linear into LTXVSaveHLG for a BT.2020/HLG master.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":true,"has_intermediate_output":false,"search_aliases":[]},"LTXVImgToVideoConditionOnly":{"input":{"required":{"vae":["VAE"],"image":["IMAGE"],"latent":["LATENT"],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0}]},"optional":{"bypass":["BOOLEAN",{"default":false,"tooltip":"Bypass the conditioning."}]}},"input_order":{"required":["vae","image","latent","strength"],"optional":["bypass"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"name":"LTXVImgToVideoConditionOnly","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Img To Video Condition Only","description":"Applies image conditioning to the first frames of an existing latent. Creates a noise mask to control conditioning strength.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"conditioning/video_models","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LTXVImgToVideoInplace":{"input":{"required":{"vae":["VAE",{}],"image":["IMAGE",{}],"latent":["LATENT",{}],"strength":["FLOAT",{"default":1.0,"min":0.0,"max":1.0}],"bypass":["BOOLEAN",{"tooltip":"Bypass the conditioning.","default":false}]}},"input_order":{"required":["vae","image","latent","strength","bypass"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVImgToVideoInplace","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt","category":"model/conditioning/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVInpaintPreprocess":{"input":{"required":{"images":["IMAGE",{"tooltip":"Video frames to composite onto the green background."}],"mask":["MASK",{"tooltip":"Mask indicating regions to replace with green. Single-frame masks are broadcast."}]}},"input_order":{"required":["images","mask"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["image"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVInpaintPreprocess","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Inpaint Preprocess","description":"Composites images with a green (#66FF00) background where mask is active, for inpainting conditioning. Always paints SDR #66FF00 (not ACEScct-converted) \u2014 that key matches IC-LoRA training even on native HDR / ACEScct plates.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/image_processing","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVLaplacianPyramidBlend":{"input":{"required":{"image_a":["IMAGE",{"tooltip":"First source image."}],"image_b":["IMAGE",{"tooltip":"Second source image."}],"mask":["MASK",{"tooltip":"Blend mask (white = image_a, black = image_b)."}],"trim_to_shortest":["BOOLEAN",{"tooltip":"Trim image_a, image_b, and mask to the shortest sequence length before blending.","default":true}],"mask_low_res_dilation":["INT",{"tooltip":"Downscale the mask to long side 64, dilate it spatially, then resize it back before blending.","default":5,"min":0,"max":15}]}},"input_order":{"required":["image_a","image_b","mask","trim_to_shortest","mask_low_res_dilation"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["image"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVLaplacianPyramidBlend","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTX Laplacian Pyramid Blend","description":"Blend two images seamlessly using Laplacian pyramid blending.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/utility","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVLatentUpsampler":{"input":{"required":{"samples":["LATENT",{}],"upscale_model":["LATENT_UPSCALE_MODEL",{}],"vae":["VAE",{}]}},"input_order":{"required":["samples","upscale_model","vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVLatentUpsampler","display_name":null,"description":"","python_module":"comfy_extras.nodes_lt_upsampler","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVLoadConditioning":{"input":{"required":{"file_name":["COMBO",{"multiselect":false,"options":[""]}],"device":["COMBO",{"multiselect":false,"options":["cpu","gpu"]}]}},"input_order":{"required":["file_name","device"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVLoadConditioning","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Load Conditioning","description":"Load precomputed text conditioning from models/embeddings. Supports Comfy conditioning_data_* files and pipeline scene embeddings (video_context / audio_context), e.g. HDR IC-LoRA.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"lightricks/LTXV","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVPreprocess":{"input":{"required":{"image":["IMAGE",{}],"img_compression":["INT",{"tooltip":"Amount of compression to apply on image.","default":35,"min":0,"max":100}]}},"input_order":{"required":["image","img_compression"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["output_image"],"output_tooltips":[null],"output_matchtypes":null,"name":"LTXVPreprocess","display_name":"LTXV Preprocess","description":"","python_module":"comfy_extras.nodes_lt","category":"video/preprocessors","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVSDRToHDRWorkingSpace":{"input":{"required":{"image":["IMAGE"],"color_space":[["srgb","srgb_gamma","acescg","acescct"],{"default":"srgb_gamma","tooltip":"SDR ingest IDT. srgb_gamma = display sRGB (EOTF then ACEScct); srgb = scene-linear Rec.709; acescg / acescct for float EXR-style plates."}]}},"input_order":{"required":["image","color_space"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["acescct"],"name":"LTXVSDRToHDRWorkingSpace","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXVSDR To HDR Working Space","description":"SDR \u2192 ACEScct working space for HDR IC-LoRA (pipelines SDRColorSpace). Wire LoadVideo images through this before resize / IC-LoRA guide.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LTXVSaveHLG":{"input":{"required":{"hdr_linear":["IMAGE",{"tooltip":"Scene-linear HDR from LTXVHDRDecodePostprocess. With transfer=acescct this is ACEScg linear \u2014 leave linear_primaries=acescg. For logc3, use rec709."}],"frame_rate":["FLOAT",{"default":24.0,"min":1.0,"max":120.0,"step":0.001}],"filename_prefix":["STRING",{"default":"hdr/ltxv_hlg"}]},"optional":{"linear_primaries":[["acescg","rec709"],{"default":"acescg","tooltip":"Primaries of hdr_linear. ACEScct decode (SDR\u2192HDR IC-LoRA or native HDR) \u2192 acescg. Legacy LogC3 \u2192 rec709."}],"audio":["AUDIO",{"tooltip":"Optional track to mux into the HLG MP4 (e.g. Decode AUDIO). Omit for video-only."}]}},"input_order":{"required":["hdr_linear","frame_rate","filename_prefix"],"optional":["linear_primaries","audio"]},"is_input_list":false,"output":[],"output_is_list":[],"output_name":[],"name":"LTXVSaveHLG","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXV Save HLG","description":"Writes a BT.2020/HLG 10-bit HEVC master from scene-linear hdr_linear (from LTXVHDRDecodePostprocess). Maps linear_primaries \u2192 Rec.2020 directly (no Rec.709 hop), then HLG. Optional AUDIO is muxed as AAC.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":true,"has_intermediate_output":false,"search_aliases":[]},"LTXVSeparateAVLatent":{"input":{"required":{"av_latent":["LATENT",{}]}},"input_order":{"required":["av_latent"]},"is_input_list":false,"output":["LATENT","LATENT"],"output_is_list":[false,false],"output_name":["video_latent","audio_latent"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"LTXVSeparateAVLatent","display_name":"Separate AV Latent","description":"Split a joint AV latent into its video and audio latents (any AV model, e.g. LTXV or MiniMax H3).","python_module":"comfy_extras.nodes_lt","category":"model/latent/ltxv","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVSetAudioRefTokens":{"input":{"required":{"positive":["CONDITIONING",{"tooltip":"Positive conditioning to attach the reference audio tokens to."}],"negative":["CONDITIONING",{"tooltip":"Negative conditioning to attach the reference audio tokens to."}],"audio_latent":["LATENT",{"tooltip":"Encoded audio latent from LTXV Audio VAE Encode."}]}},"input_order":{"required":["positive","negative","audio_latent"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","frozen_audio"],"output_tooltips":["Positive conditioning with reference audio tokens attached.","Negative conditioning with reference audio tokens attached.","Audio latent with noise_mask=0, fully frozen during denoising."],"output_matchtypes":null,"name":"LTXVSetAudioRefTokens","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 Set Audio Ref Tokens","description":"Provides speaker identity context for audio generation by attaching reference audio tokens to the conditioning. The tokens are prepended with negative temporal positions so the model treats them as context rather than generation targets.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/IC-LoRA","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LTXVVAEForceFloat32":{"input":{"required":{"vae":["VAE"]}},"input_order":{"required":["vae"]},"is_input_list":false,"output":["VAE"],"output_is_list":[false],"output_name":["VAE"],"name":"LTXVVAEForceFloat32","display_name":"\ud83c\udd5b\ud83c\udd63\ud83c\udd67 LTXVVAE Force Float32","description":"Forces float32 on the video VAE (vae_dtype + first_stage_model) for native HDR. IMPORTANT: this does NOT clone a float32 VAE for the output \u2014 it mutates the input object in place and returns the same instance. If you also wire that loader's VAE (or this output) into another branch, those nodes share the mutated model: they may see float16 or float32 depending on execution order (a race with the original dtype). Prefer a single chain VAELoader \u2192 this node \u2192 all encode/decode / IC-LoRA guide uses.","python_module":"custom_nodes.ComfyUI-LTXVideo","category":"Lightricks/HDR","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"LatentUpscaleModelLoader":{"input":{"required":{"model_name":["COMBO",{"multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["model_name"]},"is_input_list":false,"output":["LATENT_UPSCALE_MODEL"],"output_is_list":[false],"output_name":["LATENT_UPSCALE_MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"LatentUpscaleModelLoader","display_name":"Load Latent Upscale Model","description":"","python_module":"comfy_extras.nodes_hunyuan","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LoadAudio":{"input":{"required":{"audio":["COMBO",{"multiselect":false,"options":["clip.mp4","drive_doll_street_dance.mp4","grainy_perfume_shot_crf32.mp4","ltx_23_audio.mp3","stone_ruins.mp4"],"audio_upload":true}]}},"input_order":{"required":["audio"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadAudio","display_name":"Load Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["import audio","open audio","audio file"],"essentials_category":"Audio","has_intermediate_output":false},"LoadBackgroundRemovalModel":{"input":{"required":{"bg_removal_name":["COMBO",{"tooltip":"The model used to remove backgrounds from images","multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["bg_removal_name"]},"is_input_list":false,"output":["BACKGROUND_REMOVAL"],"output_is_list":[false],"output_name":["bg_model"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadBackgroundRemovalModel","display_name":"Load Background Removal Model","description":"","python_module":"comfy_extras.nodes_bg_removal","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LoadImage":{"input":{"required":{"image":[["cactus_man.png","example.png","fashion_portrait_blue_silk_gown_vis.png","lab_smoke-invert-chain_smoke_00005_.png","ltx23_flf2v_first_frame.png","ltx23_flf2v_last_frame.png","neon_cyborg_portrait.png","neon_guitarist.png","portrait.png","ref_doll_realistic.png","robot_hand.png","robot_hand_energy.png","server_out.png","smoke_00005_.png","the_forgotten_gate.png","the_lily_veil.png","watch_macro_shot.png","woman_gold_earring_teal_top_portrait.png"],{"image_upload":true}]}},"input_order":{"required":["image"]},"is_input_list":false,"output":["IMAGE","MASK"],"output_is_list":[false,false],"output_name":["IMAGE","MASK"],"name":"LoadImage","display_name":"Load Image","description":"","python_module":"nodes","category":"image","output_node":false,"has_intermediate_output":false,"search_aliases":["load image","open image","import image","image input","upload image","read image","image loader"],"essentials_category":"Basics"},"LoadMoGeModel":{"input":{"required":{"model_name":["COMBO",{"multiselect":false,"options":["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]}]}},"input_order":{"required":["model_name"]},"is_input_list":false,"output":["MOGE_MODEL"],"output_is_list":[false],"output_name":["MOGE_MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadMoGeModel","display_name":"Load MoGe Model","description":"","python_module":"comfy_extras.nodes_moge","category":"model/loaders","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"LoadVideo":{"input":{"required":{"file":["COMBO",{"multiselect":false,"options":["clip.mp4","drive_doll_street_dance.mp4","grainy_perfume_shot_crf32.mp4","stone_ruins.mp4"],"video_upload":true}]}},"input_order":{"required":["file"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["VIDEO"],"output_tooltips":[null],"output_matchtypes":null,"name":"LoadVideo","display_name":"Load Video","description":"","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["import video","open video","video file"],"essentials_category":"Basics","has_intermediate_output":true},"LoraLoaderModelOnly":{"input":{"required":{"model":["MODEL"],"lora_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]],"strength_model":["FLOAT",{"default":1.0,"min":-100.0,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","lora_name","strength_model"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"LoraLoaderModelOnly","display_name":"Load LoRA","description":"This LoRAs loader is used to modify the diffusion model, altering the way in which latents are denoised such as applying styles. Multiple LoRA nodes can be linked together.","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"output_tooltips":["The modified diffusion model.","The modified CLIP model."],"search_aliases":["lora","load lora","apply lora","lora loader","lora model"],"essentials_category":"Image Generation"},"ManualSigmas":{"input":{"required":{"sigmas":["STRING",{"default":"1, 0.5","multiline":false}]}},"input_order":{"required":["sigmas"]},"is_input_list":false,"output":["SIGMAS"],"output_is_list":[false],"output_name":["SIGMAS"],"output_tooltips":[null],"output_matchtypes":null,"name":"ManualSigmas","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/sigmas","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["custom noise schedule","define sigmas"],"essentials_category":null,"has_intermediate_output":false},"MaskPreview":{"input":{"required":{"mask":["MASK",{}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["mask"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["mask"],"output_tooltips":[null],"output_matchtypes":null,"name":"MaskPreview","display_name":"Preview Mask","description":"Preview the masks without saving them to the ComfyUI output directory.","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["preview","preview mask","show mask","view mask","inspect mask","debug mask"],"essentials_category":null,"has_intermediate_output":false},"MaskToImage":{"input":{"required":{"mask":["MASK",{}]}},"input_order":{"required":["mask"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"MaskToImage","display_name":"Convert Mask to Image","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["convert mask"],"essentials_category":null,"has_intermediate_output":false},"MiniMaxH3AddGuide":{"input":{"required":{"positive":["CONDITIONING",{}],"latent":["LATENT",{}],"frame_idx":["INT",{"tooltip":"Frame index to anchor the image or the clip's first frame at. Negative values are counted from the end of the video.","default":0,"min":-9999,"max":9999}]},"optional":{"vae":["VAE",{"tooltip":"Video VAE, needed when an image is connected."}],"audio_vae":["VAE",{"tooltip":"Audio VAE, needed when an audio is connected."}],"image":["IMAGE",{"tooltip":"Image or video frames to anchor. Multi-frame batches are anchored as a clip and cropped down to the model's valid clip lengths: 5, 22, 39... (17k + 5) frames. Batches shorter than 5 frames use only the first image."}],"audio":["AUDIO",{"tooltip":"Soundtrack to anchor starting at the same frame index, cropped to the video's remaining duration."}]}},"input_order":{"required":["positive","latent","frame_idx"],"optional":["vae","audio_vae","image","audio"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["positive"],"output_tooltips":[null],"output_matchtypes":null,"name":"MiniMaxH3AddGuide","display_name":"Add Guide for MiniMax H3","description":"Anchor an image, a short clip, audio, or a clip with its soundtrack at any frame of a MiniMax H3 video. Chain several nodes to anchor several frames.","python_module":"comfy_extras.nodes_minimax_h3","category":"model/conditioning/minimax","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"MiniMaxH3ImageToVideo":{"input":{"required":{"clip":["CLIP",{}],"vae":["VAE",{}],"prompt":["STRING",{"multiline":true,"dynamicPrompts":true}],"width":["INT",{"default":1344,"min":32,"max":16384,"step":32}],"height":["INT",{"default":768,"min":32,"max":16384,"step":32}],"length":["INT",{"tooltip":"Frame count at 24 fps, snapped up to the model's 17k+5 grid (124 = ~5s; trained range is ~124-362, longer is untested)","default":124,"min":5,"max":3600,"step":17}]},"optional":{"first_frame":["IMAGE",{}],"last_frame":["IMAGE",{}]}},"input_order":{"required":["clip","vae","prompt","width","height","length"],"optional":["first_frame","last_frame"]},"is_input_list":false,"output":["CONDITIONING","LATENT"],"output_is_list":[false,false],"output_name":["positive","LATENT"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"MiniMaxH3ImageToVideo","display_name":"MiniMax H3 Image to Video","description":"","python_module":"comfy_extras.nodes_minimax_h3","category":"model/conditioning/minimax","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"MiniMaxH3ReferenceToVideo":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"multiline":true,"dynamicPrompts":true}],"width":["INT",{"default":1344,"min":32,"max":16384,"step":32}],"height":["INT",{"default":768,"min":32,"max":16384,"step":32}],"length":["INT",{"tooltip":"Frame count at 24 fps, (124 = ~5s, trained range is ~124-362)","default":124,"min":5,"max":3600,"step":17}],"ref_image_size":["COMBO",{"tooltip":"Reference image sizing. 'match' scales each ref (down only, keeping aspect) to the generation's pixel area; 'max' uses the reference pipeline's 2048px short edge for best identity fidelity. Reference tokens ride through every sampling step, so 'max' can be several times slower.","default":"match","multiselect":false,"options":["match","max"]}]},"optional":{"vae":["VAE",{"tooltip":"Video VAE. Without it reference images/videos only condition the text encoder."}],"audio_vae":["VAE",{"tooltip":"Audio VAE. Without it reference audio only conditions the text encoder."}],"ref_images":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_image":["IMAGE",{"tooltip":"Reference image (downscaled to 2048 short edge if larger, never upscaled)"}]}},"prefix":"ref_image_","min":0,"max":9}}],"ref_videos":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_video":["IMAGE",{"tooltip":"Reference video frames at 24 fps (2-15s)"}]}},"prefix":"ref_video_","min":0,"max":3}}],"ref_video_audios":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_video_audio":["AUDIO",{"tooltip":"Soundtrack of the same-numbered reference video"}]}},"prefix":"ref_video_audio_","min":0,"max":3}}],"ref_audios":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{"ref_audio":["AUDIO",{"tooltip":"Standalone reference audio"}]}},"prefix":"ref_audio_","min":0,"max":3}}]}},"input_order":{"required":["clip","prompt","width","height","length","ref_image_size"],"optional":["vae","audio_vae","ref_images","ref_videos","ref_video_audios","ref_audios"]},"is_input_list":false,"output":["CONDITIONING","LATENT"],"output_is_list":[false,false],"output_name":["positive","LATENT"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"MiniMaxH3ReferenceToVideo","display_name":"MiniMax H3 Reference to Video","description":"<Picture i> / <Video k> / <Audio j> reference conditioning for MiniMax H3. Use the same tags when prompting.","python_module":"comfy_extras.nodes_minimax_h3","category":"model/conditioning/minimax","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"MiniMaxH3SigmaShift":{"input":{"required":{"model":["MODEL",{}],"shift_video":["FLOAT",{"default":12.0,"min":0.01,"max":100.0,"step":0.01}],"shift_audio":["FLOAT",{"default":3.0,"min":0.01,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","shift_video","shift_audio"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"MiniMaxH3SigmaShift","display_name":"ModelSamplingMiniMaxH3","description":"Set the video/audio flow shifts.","python_module":"comfy_extras.nodes_minimax_h3","category":"model/patch/minimax","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["sigma shift","minimax shift"],"essentials_category":null,"has_intermediate_output":false},"MiniMaxMusic3TextEncode":{"input":{"required":{"clip":["CLIP",{}],"caption":["STRING",{"multiline":true,"dynamicPrompts":true}],"lyrics":["STRING",{"multiline":true,"dynamicPrompts":true}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}],"max_duration":["FLOAT",{"tooltip":"Maximum duration in seconds; the model can end the song earlier.","default":120.0,"min":0.04,"max":360.0,"step":0.04}],"cfg_scale":["FLOAT",{"advanced":true,"default":1.5,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"top_k":["INT",{"advanced":true,"default":50,"min":1,"max":16384}]}},"input_order":{"required":["clip","caption","lyrics","seed","max_duration","cfg_scale","top_k"]},"is_input_list":false,"output":["CONDITIONING","FLOAT"],"output_is_list":[false,false],"output_name":["CONDITIONING","seconds"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"MiniMaxMusic3TextEncode","display_name":"MiniMax Music3 Text Encode","description":"Uses a MiniMax Music3 CLIP model to generate the acoustic conditioning sequence.","python_module":"comfy_extras.nodes_minimax_music","category":"model/conditioning/minimax music","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"MoGeInference":{"input":{"required":{"moge_model":["MOGE_MODEL",{}],"image":["IMAGE",{}],"resolution_level":["INT",{"tooltip":"0 = fastest, 9 = most detail.","default":9,"min":0,"max":9}],"fov_x_degrees":["FLOAT",{"tooltip":"Horizontal field of view of the source camera. Sets the focal length used to unproject the depth map into 3D. 0 = auto-recover from the predicted points.","advanced":true,"default":0.0,"min":0.0,"max":170.0,"step":0.1}],"batch_size":["INT",{"tooltip":"Images per inference call. Lower if you OOM on a long video / image set.","default":4,"min":1,"max":64}],"force_projection":["BOOLEAN",{"advanced":true,"default":true}],"apply_mask":["BOOLEAN",{"tooltip":"Set masked-out (sky / invalid) pixels to inf in points and depth so meshing culls them. Disable to keep the raw predicted geometry everywhere; the mask is still returned separately.","advanced":true,"default":true}],"refine_steps":["INT",{"tooltip":"MoGe-3 only: sparse volumetric refinement passes over the predicted depth. More passes sharpen fine detail and edges at a roughly linear cost. 0 disables refinement. Ignored by MoGe-1 / MoGe-2.","advanced":true,"default":3,"min":0,"max":8}]}},"input_order":{"required":["moge_model","image","resolution_level","fov_x_degrees","batch_size","force_projection","apply_mask","refine_steps"]},"is_input_list":false,"output":["MOGE_GEOMETRY"],"output_is_list":[false],"output_name":["moge_geometry"],"output_tooltips":[null],"output_matchtypes":null,"name":"MoGeInference","display_name":"Run MoGe Inference","description":"Run MoGe on a single image to estimate depth and geometry.","python_module":"comfy_extras.nodes_moge","category":"image/geometry estimation","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["moge","depth","geometry","depth estimation","geometry estimation"],"essentials_category":null,"has_intermediate_output":false},"MoGeRender":{"input":{"required":{"moge_geometry":["MOGE_GEOMETRY",{}],"output":["COMBO",{"tooltip":"DirectX vs OpenGL controls the normal-map green-channel convention. DirectX: green = -Y down (Unreal). OpenGL: green = +Y up (Blender, Substance, Unity, glTF).","default":"depth","multiselect":false,"options":["depth","depth_colored","normal_opengl","normal_directx","mask"]}]}},"input_order":{"required":["moge_geometry","output"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"MoGeRender","display_name":"Render MoGe Geometry","description":"Render a depth map or normal map from geometry data","python_module":"comfy_extras.nodes_moge","category":"image/geometry estimation","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["moge","render","geometry","depth","normal"],"essentials_category":null,"has_intermediate_output":false},"ModelAttentionBackend":{"input":{"required":{"model":["MODEL",{"tooltip":"The model to patch."}],"attention":["COMBO",{"display_name":"backend","tooltip":"The dense attention backend. Comfy Kitchen attention uses quantized INT8 attention and is available only on Nvidia and AMD GPUs.","default":"pytorch attention","multiselect":false,"options":["pytorch attention"]}]}},"input_order":{"required":["model","attention"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["model"],"output_tooltips":["The model with the selected attention backend."],"output_matchtypes":null,"name":"ModelAttentionBackend","display_name":"Model Attention Backend","description":"Selects the dense attention implementation for the model. When used with Block Sparse Attention, this backend is used whenever sparse attention is inactive or unsupported.","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["comfy kitchen","attention","pytorch","kitchen attention"],"essentials_category":null,"has_intermediate_output":false},"ModelSamplingAuraFlow":{"input":{"required":{"model":["MODEL"],"shift":["FLOAT",{"default":1.73,"min":0.0,"max":100.0,"step":0.01}]},"optional":{"sampling":[["flow","img_to_img_velocity"],{"default":"flow","advanced":true}]}},"input_order":{"required":["model","shift"],"optional":["sampling"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"ModelSamplingAuraFlow","display_name":"ModelSamplingAuraFlow","description":"","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"ModelSamplingFlux":{"input":{"required":{"model":["MODEL"],"max_shift":["FLOAT",{"default":1.15,"min":0.0,"max":100.0,"step":0.01,"advanced":true}],"base_shift":["FLOAT",{"default":0.5,"min":0.0,"max":100.0,"step":0.01,"advanced":true}],"width":["INT",{"default":1024,"min":16,"max":16384,"step":8}],"height":["INT",{"default":1024,"min":16,"max":16384,"step":8}]}},"input_order":{"required":["model","max_shift","base_shift","width","height"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"ModelSamplingFlux","display_name":"ModelSamplingFlux","description":"","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch/flux","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"ModelSamplingSD3":{"input":{"required":{"model":["MODEL"],"shift":["FLOAT",{"default":3.0,"min":0.0,"max":100.0,"step":0.01}]}},"input_order":{"required":["model","shift"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"ModelSamplingSD3","display_name":"ModelSamplingSD3","description":"","python_module":"comfy_extras.nodes_model_advanced","category":"model/patch/stable diffusion","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"PreviewAny":{"input":{"required":{"source":["*",{}]}},"input_order":{"required":["source"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"name":"PreviewAny","display_name":"Preview as Text","description":"Preview any input value as text.","python_module":"comfy_extras.nodes_preview_any","category":"utilities","output_node":true,"has_intermediate_output":false,"search_aliases":["preview","preview text","show output","inspect","debug","print value","show text"]},"PreviewImage":{"input":{"required":{"images":["IMAGE"]},"hidden":{"prompt":"PROMPT","extra_pnginfo":"EXTRA_PNGINFO"}},"input_order":{"required":["images"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"name":"PreviewImage","display_name":"Preview Image","description":"Preview the images without saving them to the ComfyUI output directory.","python_module":"nodes","category":"image","output_node":true,"has_intermediate_output":false,"search_aliases":["preview","preview image","show image","view image","display image","image viewer"],"essentials_category":"Basics"},"PrimitiveBoolean":{"input":{"required":{"value":["BOOLEAN",{}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["BOOLEAN"],"output_is_list":[false],"output_name":["BOOLEAN"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveBoolean","display_name":"Boolean","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"PrimitiveFloat":{"input":{"required":{"value":["FLOAT",{"min":-9223372036854775807,"max":9223372036854775807,"step":0.1}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["FLOAT"],"output_is_list":[false],"output_name":["FLOAT"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveFloat","display_name":"Float","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"PrimitiveInt":{"input":{"required":{"value":["INT",{"min":-9223372036854775807,"max":9223372036854775807,"control_after_generate":"fixed"}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["INT"],"output_is_list":[false],"output_name":["INT"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveInt","display_name":"Int","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"PrimitiveString":{"input":{"required":{"value":["STRING",{"multiline":false}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveString","display_name":"Text","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["text","string","text box","prompt"],"essentials_category":null,"has_intermediate_output":false},"PrimitiveStringMultiline":{"input":{"required":{"value":["STRING",{"multiline":true}]}},"input_order":{"required":["value"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"PrimitiveStringMultiline","display_name":"Text (Multiline)","description":"","python_module":"comfy_extras.nodes_primitive","category":"utilities/primitive","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["text","string","text box","prompt"],"essentials_category":"Basics","has_intermediate_output":false},"RandomNoise":{"input":{"required":{"noise_seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}]}},"input_order":{"required":["noise_seed"]},"is_input_list":false,"output":["NOISE"],"output_is_list":[false],"output_name":["NOISE"],"output_tooltips":[null],"output_matchtypes":null,"name":"RandomNoise","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/noise","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"ReferenceMultiLatent":{"input":{"required":{"positive":["CONDITIONING",{}],"resolution":["COMFY_DYNAMICCOMBO_V3",{"options":[{"key":"keep resolution","inputs":{"required":{}}},{"key":"resize","inputs":{"required":{"megapixels":["FLOAT",{"tooltip":"Target size of each reference image, preserving aspect ratio. 1.0 is about 1024 x 1024 pixels.","default":1.0,"min":0.01,"max":16.0,"step":0.01}]}}}]}]},"optional":{"negative":["CONDITIONING",{}],"vae":["VAE",{"tooltip":"Reference images are ignored when no VAE is connected."}],"images":["COMFY_AUTOGROW_V3",{"template":{"input":{"required":{},"optional":{"image":["IMAGE",{}]}},"prefix":"image","min":1,"max":100}}]}},"input_order":{"required":["positive","resolution"],"optional":["negative","vae","images"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","empty_latent"],"output_tooltips":[null,"Returns the positive output when negative conditioning is not connected.","First reference latent multiplied by 0.0, or a square empty latent at the selected resolution."],"output_matchtypes":null,"name":"ReferenceMultiLatent","display_name":"Set Reference Multi Latent","description":"Encodes reference images and appends their latents to positive and optional negative conditioning for edit models.","python_module":"comfy_extras.nodes_edit_model","category":"model/conditioning","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"RemoveBackground":{"input":{"required":{"bg_removal_model":["BACKGROUND_REMOVAL",{"tooltip":"Background removal model used to generate the mask"}],"image":["IMAGE",{"tooltip":"Input image to remove the background from"}]}},"input_order":{"required":["bg_removal_model","image"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["mask"],"output_tooltips":["Generated foreground mask"],"output_matchtypes":null,"name":"RemoveBackground","display_name":"Remove Background","description":"Generates a foreground mask to remove the background from an image using a background removal model.","python_module":"comfy_extras.nodes_bg_removal","category":"image/background removal","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"RepeatImageBatch":{"input":{"required":{"image":["IMAGE",{}],"amount":["INT",{"default":1,"min":1,"max":4096}]}},"input_order":{"required":["image","amount"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"output_tooltips":[null],"output_matchtypes":null,"name":"RepeatImageBatch","display_name":"Repeat Image Batch","description":"","python_module":"comfy_extras.nodes_images","category":"image/batch","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["duplicate image","clone image"],"essentials_category":null,"has_intermediate_output":false},"ResizeImageMaskNode":{"input":{"required":{"input":["COMFY_MATCHTYPE_V3",{"template":{"template_id":"input_type","allowed_types":"IMAGE,MASK"}}],"resize_type":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"Select how to resize: by exact dimensions, scale factor, matching another image, etc.","options":[{"key":"scale dimensions","inputs":{"required":{"width":["INT",{"tooltip":"Target width in pixels. Set to 0 to auto-calculate from height while preserving aspect ratio.","default":512,"min":0,"max":16384,"step":1}],"height":["INT",{"tooltip":"Target height in pixels. Set to 0 to auto-calculate from width while preserving aspect ratio.","default":512,"min":0,"max":16384,"step":1}],"crop":["COMBO",{"tooltip":"How to handle aspect ratio mismatch: 'disabled' stretches to fit, 'center' crops to maintain aspect ratio.","default":"center","multiselect":false,"options":["disabled","center"]}]}}},{"key":"scale by multiplier","inputs":{"required":{"multiplier":["FLOAT",{"tooltip":"Scale factor (e.g., 2.0 doubles size, 0.5 halves size).","default":1.0,"min":0.01,"max":8.0,"step":0.01}]}}},{"key":"scale longer dimension","inputs":{"required":{"longer_size":["INT",{"tooltip":"The longer edge will be resized to this value. Aspect ratio is preserved.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale shorter dimension","inputs":{"required":{"shorter_size":["INT",{"tooltip":"The shorter edge will be resized to this value. Aspect ratio is preserved.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale width","inputs":{"required":{"width":["INT",{"tooltip":"Target width in pixels. Height auto-adjusts to preserve aspect ratio.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale height","inputs":{"required":{"height":["INT",{"tooltip":"Target height in pixels. Width auto-adjusts to preserve aspect ratio.","default":512,"min":0,"max":16384,"step":1}]}}},{"key":"scale total pixels","inputs":{"required":{"megapixels":["FLOAT",{"tooltip":"Target total megapixels (e.g., 1.0 \u2248 1024\u00d71024). Aspect ratio is preserved.","default":1.0,"min":0.01,"max":16.0,"step":0.01}]}}},{"key":"match size","inputs":{"required":{"match":["IMAGE,MASK",{"tooltip":"Resize input to match the dimensions of this reference image or mask."}],"crop":["COMBO",{"tooltip":"How to handle aspect ratio mismatch: 'disabled' stretches to fit, 'center' crops to maintain aspect ratio.","default":"center","multiselect":false,"options":["disabled","center"]}]}}},{"key":"scale to multiple","inputs":{"required":{"multiple":["INT",{"tooltip":"Resize so width and height are divisible by this number. Useful for latent alignment (e.g., 8 or 64).","default":8,"min":1,"max":16384,"step":1}]}}}]}],"scale_method":["COMBO",{"tooltip":"Interpolation algorithm. 'area' is best for downscaling, 'lanczos' for upscaling, 'nearest-exact' for pixel art.","default":"area","multiselect":false,"options":["nearest-exact","bilinear","area","bicubic","lanczos"]}]}},"input_order":{"required":["input","resize_type","scale_method"]},"is_input_list":false,"output":["COMFY_MATCHTYPE_V3"],"output_is_list":[false],"output_name":["resized"],"output_tooltips":[null],"output_matchtypes":["input_type"],"name":"ResizeImageMaskNode","display_name":"Resize Image/Mask","description":"Resize an image or mask using various scaling methods.","python_module":"comfy_extras.nodes_post_processing","category":"image/transform","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["resize","resize image","resize mask","scale","scale image","scale mask","image resize","change size","dimensions","shrink","enlarge"],"essentials_category":null,"has_intermediate_output":false},"ResolutionSelector":{"input":{"required":{"aspect_ratio":["COMBO",{"tooltip":"The aspect ratio for the output dimensions.","default":"1:1 (Square)","multiselect":false,"options":["1:1 (Square)","2:3 (Portrait Photo)","3:2 (Photo)","3:4 (Portrait Standard)","4:3 (Standard)","9:16 (Portrait Widescreen)","16:9 (Widescreen)","21:9 (Ultrawide)"]}],"megapixels":["FLOAT",{"tooltip":"Target total megapixels. 1.0 MP \u2248 1024x1024 for square.","default":1.0,"min":0.1,"max":16.0,"step":0.1}],"multiple":["INT",{"tooltip":"Nearest multiple of the result to set the selected resolution to.","advanced":true,"default":8,"min":8,"max":128,"step":4}]},"optional":{"preview":["RESOLUTION_PREVIEW",{"tooltip":"Live preview of the calculated output resolution.","socketless":true,"ratio_widget":"aspect_ratio","megapixels_widget":"megapixels","multiple_widget":"multiple"}]}},"input_order":{"required":["aspect_ratio","megapixels","multiple"],"optional":["preview"]},"is_input_list":false,"output":["INT","INT"],"output_is_list":[false,false],"output_name":["width","height"],"output_tooltips":["Calculated width in pixels multiplied by the selected multiple.","Calculated height in pixels multiplied by the selected multiple."],"output_matchtypes":null,"name":"ResolutionSelector","display_name":"Resolution Selector","description":"Calculate width and height from aspect ratio and megapixel target. Useful for setting up Empty Latent Image dimensions.","python_module":"comfy_extras.nodes_resolution","category":"utilities","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SAM3_VideoTrack":{"input":{"required":{"images":["IMAGE",{"display_name":"images","tooltip":"Video frames as batched images"}],"model":["MODEL",{"display_name":"model"}],"detection_threshold":["FLOAT",{"display_name":"detection_threshold","tooltip":"Score threshold for text-prompted detection.","default":0.5,"min":0.0,"max":1.0,"step":0.01}],"max_objects":["INT",{"display_name":"max_objects","tooltip":"Max tracked objects. Initial masks count toward this limit. 0 uses the internal cap of 64.","default":4,"min":0,"max":64}],"detect_interval":["INT",{"display_name":"detect_interval","tooltip":"Run detection every N frames (1=every frame). Higher values save compute.","default":1,"min":1}]},"optional":{"initial_mask":["MASK",{"display_name":"initial_mask","tooltip":"Mask(s) for the first frame to track (one per object)"}],"conditioning":["CONDITIONING",{"display_name":"conditioning","tooltip":"Text conditioning for detecting new objects during tracking"}]}},"input_order":{"required":["images","model","detection_threshold","max_objects","detect_interval"],"optional":["initial_mask","conditioning"]},"is_input_list":false,"output":["SAM3_TRACK_DATA"],"output_is_list":[false],"output_name":["track_data"],"output_tooltips":[null],"output_matchtypes":null,"name":"SAM3_VideoTrack","display_name":"Run SAM3 Video Track","description":"","python_module":"comfy_extras.nodes_sam3","category":"image/detection","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["sam3","video","track","propagate"],"essentials_category":null,"has_intermediate_output":false},"SCAIL2ColoredMask":{"input":{"required":{"driving_track_data":["SAM3_TRACK_DATA",{"tooltip":"SAM3 track of the driving pose video. Will be rendered into the pose_video_mask output."}],"object_indices":["STRING",{"tooltip":"Comma-separated list of person indices to include (e.g. '0,2,3'). Applied to both reference and pose video masks. Empty = all.","default":"","multiline":false}],"sort_by":["COMBO",{"tooltip":"Order in which palette colors are assigned to the tracked objects (applied to both reference and pose video so each identity keeps the same color). Objects that appear in earlier frames always come first; within a frame, left_to_right = leftmost object (by centroid at first appearance) gets the first color, area = biggest object (by mask area at first appearance) gets the first color; none = keep SAM3's order.","default":"left_to_right","multiselect":false,"options":["none","left_to_right","area"]}],"replacement_mode":["BOOLEAN",{"tooltip":"False = Animation Mode (pose_video_mask has black background, reference_image_mask has white background). True = Replacement Mode (pose_video_mask has white background, reference_image_mask has black background).","default":false}]},"optional":{"ref_track_data":["SAM3_TRACK_DATA,MASK",{"display_name":"reference_masks","tooltip":"SAM3 track of the reference image(s) (one identity per object, colored in batch order), or a plain MASK of the reference subject (rendered as a single identity)."}]}},"input_order":{"required":["driving_track_data","object_indices","sort_by","replacement_mode"],"optional":["ref_track_data"]},"is_input_list":false,"output":["IMAGE","IMAGE"],"output_is_list":[false,false],"output_name":["pose_video_mask","reference_image_mask"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"SCAIL2ColoredMask","display_name":"Create SCAIL-2 Colored Mask","description":"","python_module":"comfy_extras.nodes_scail","category":"model/conditioning/wan/scail","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SamplerCustom":{"input":{"required":{"model":["MODEL",{}],"add_noise":["BOOLEAN",{"advanced":true,"default":true}],"noise_seed":["INT",{"default":0,"min":0,"max":18446744073709551615,"control_after_generate":true}],"cfg":["FLOAT",{"default":8.0,"min":0.0,"max":100.0,"step":0.1,"round":0.01}],"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"sampler":["SAMPLER",{}],"sigmas":["SIGMAS",{}],"latent_image":["LATENT",{}]}},"input_order":{"required":["model","add_noise","noise_seed","cfg","positive","negative","sampler","sigmas","latent_image"]},"is_input_list":false,"output":["LATENT","LATENT"],"output_is_list":[false,false],"output_name":["output","denoised_output"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"SamplerCustom","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/custom","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SamplerCustomAdvanced":{"input":{"required":{"noise":["NOISE",{}],"guider":["GUIDER",{}],"sampler":["SAMPLER",{}],"sigmas":["SIGMAS",{}],"latent_image":["LATENT",{}]}},"input_order":{"required":["noise","guider","sampler","sigmas","latent_image"]},"is_input_list":false,"output":["LATENT","LATENT"],"output_is_list":[false,false],"output_name":["output","denoised_output"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"SamplerCustomAdvanced","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/custom","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SamplerEulerAncestral":{"input":{"required":{"eta":["FLOAT",{"advanced":true,"default":1.0,"min":0.0,"max":100.0,"step":0.01,"round":false}],"s_noise":["FLOAT",{"advanced":true,"default":1.0,"min":0.0,"max":100.0,"step":0.01,"round":false}]}},"input_order":{"required":["eta","s_noise"]},"is_input_list":false,"output":["SAMPLER"],"output_is_list":[false],"output_name":["SAMPLER"],"output_tooltips":[null],"output_matchtypes":null,"name":"SamplerEulerAncestral","display_name":null,"description":"","python_module":"comfy_extras.nodes_custom_sampler","category":"model/sampling/samplers","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"SaveAudioAdvanced":{"input":{"required":{"audio":["AUDIO",{"tooltip":"The audio to save."}],"filename_prefix":["STRING",{"tooltip":"The prefix for the file to save. May include formatting tokens such as %date:yyyy-MM-dd%.","default":"audio/ComfyUI","multiline":false}],"format":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The file format in which to save the audio.","options":[{"key":"flac","inputs":{"required":{}}},{"key":"mp3","inputs":{"required":{"quality":["COMBO",{"default":"V0","multiselect":false,"options":["V0","128k","320k"]}]}}},{"key":"opus","inputs":{"required":{"quality":["COMBO",{"default":"128k","multiselect":false,"options":["64k","96k","128k","192k","320k"]}]}}}]}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["audio","filename_prefix","format"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["audio"],"output_tooltips":[null],"output_matchtypes":null,"name":"SaveAudioAdvanced","display_name":"Save Audio (Advanced)","description":"Saves the input audio to your ComfyUI output directory.","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["save audio","export audio","output audio","write audio","flac","mp3","opus"],"essentials_category":null,"has_intermediate_output":false},"SaveImage":{"input":{"required":{"images":["IMAGE",{"tooltip":"The images to save."}],"filename_prefix":["STRING",{"default":"ComfyUI","tooltip":"The prefix for the file to save. This may include formatting information such as %date:yyyy-MM-dd% or %Empty Latent Image.width% to include values from nodes."}]},"hidden":{"prompt":"PROMPT","extra_pnginfo":"EXTRA_PNGINFO"}},"input_order":{"required":["images","filename_prefix"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"name":"SaveImage","display_name":"Save Image","description":"Saves the input images to your ComfyUI output directory.","python_module":"nodes","category":"image","output_node":true,"has_intermediate_output":false,"search_aliases":["save","save image","export image","output image","write image","download"],"essentials_category":"Basics"},"SaveVideo":{"input":{"required":{"video":["VIDEO",{"tooltip":"The video to save."}],"filename_prefix":["STRING",{"tooltip":"The prefix for the file to save. This may include formatting information such as %date:yyyy-MM-dd% or %Empty Latent Image.width% to include values from nodes.","default":"video/ComfyUI","multiline":false}],"format":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output container. Auto uses MP4 for Auto/H.264 and WebM for AV1. MP4, MKV, and WebM select a specific container.","options":[{"key":"auto","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}},{"key":"mp4","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}},{"key":"mkv","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}},{"key":"webm","inputs":{"required":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]}}}]}]},"optional":{"codec":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"The output video codec. Auto preserves a compatible source stream. H.264 and AV1 re-encoding support SDR, HDR (HLG), and HDR PQ.","hidden":true,"options":[{"key":"auto","inputs":{"required":{}}},{"key":"h264","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible H.264 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":18.0,"min":0.0,"max":51.0,"step":1.0}]}}}]}]}}},{"key":"av1","inputs":{"required":{},"optional":{"encoding":["COMFY_DYNAMICCOMBO_V3",{"display_name":"encoding mode","tooltip":"Automatic preserves compatible AV1 streams. Re-encode applies custom encoding options.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"re-encode","inputs":{"required":{"crf":["FLOAT",{"tooltip":"Lower values produce higher quality and larger files.","default":24.0,"min":0.0,"max":63.0,"step":1.0}]}}}]}]}}}]}]},"hidden":{"prompt":["PROMPT"],"extra_pnginfo":["EXTRA_PNGINFO"]}},"input_order":{"required":["video","filename_prefix","format"],"optional":["codec"],"hidden":["prompt","extra_pnginfo"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["video"],"output_tooltips":["The input video, unchanged."],"output_matchtypes":null,"name":"SaveVideo","display_name":"Save Video","description":"Saves the input videos to your ComfyUI output directory.","python_module":"comfy_extras.nodes_video","category":"video","output_node":true,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["export video"],"essentials_category":"Basics","has_intermediate_output":false},"SeedNode":{"input":{"required":{"seed":["INT",{"min":0,"max":9223372036854775807,"control_after_generate":"fixed"}]}},"input_order":{"required":["seed"]},"is_input_list":false,"output":["INT"],"output_is_list":[false],"output_name":["seed"],"output_tooltips":[null],"output_matchtypes":null,"name":"SeedNode","display_name":"Seed","description":"","python_module":"comfy_extras.nodes_seed","category":"utilities","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seed","random"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2Conditioning":{"input":{"required":{"model":["MODEL",{"tooltip":"The SeedVR2 model."}],"vae_conditioning":["LATENT",{"display_name":"latent"}]}},"input_order":{"required":["model","vae_conditioning"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING"],"output_is_list":[false,false],"output_name":["positive","negative"],"output_tooltips":["The positive conditioning for sampling.","The negative conditioning for sampling."],"output_matchtypes":null,"name":"SeedVR2Conditioning","display_name":"Apply SeedVR2 Conditioning","description":"Build SeedVR2 positive/negative conditioning from a VAE latent.","python_module":"comfy_extras.nodes_seedvr","category":"model/conditioning","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","upscale","conditioning"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2PostProcessing":{"input":{"required":{"images":["IMAGE",{"tooltip":"The generated image to process."}],"original_resized_images":["IMAGE",{"tooltip":"The original resized image before pre-processing, used as reference."}],"color_correction_method":["COMBO",{"tooltip":"Method to match the generated image colors to the original image. lab: transfer color in CIELAB space, preserving detail (most faithful). wavelet: transfer low-frequency color, keeping upscaled high-frequency detail. adain: match per-channel mean/std (fastest, global tint). none: skip color transfer (geometry alignment only).","default":"lab","multiselect":false,"options":["lab","wavelet","adain","none"]}]}},"input_order":{"required":["images","original_resized_images","color_correction_method"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"output_tooltips":["The aligned, color-corrected image."],"output_matchtypes":null,"name":"SeedVR2PostProcessing","display_name":"Post-Process SeedVR2 Output","description":"Align the generated image with the original resized image and apply color correction.","python_module":"comfy_extras.nodes_seedvr","category":"image/post-processors","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","upscale","color correction","color match","postprocess"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2Preprocess":{"input":{"required":{"resized_images":["IMAGE",{"tooltip":"The resized image to process."}]}},"input_order":{"required":["resized_images"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["images"],"output_tooltips":["The padded image for VAE encoding."],"output_matchtypes":null,"name":"SeedVR2Preprocess","display_name":"Pre-Process SeedVR2 Input","description":"Pad a resized image for SeedVR2 model. Alpha channel is dropped. The node Post-Process SeedVR2 Output re-applies it from the original resized image.","python_module":"comfy_extras.nodes_seedvr","category":"image/pre-processors","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","upscale","video upscale","pad","preprocess"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2TemporalChunk":{"input":{"required":{"latent":["LATENT",{"tooltip":"The VAE-encoded SeedVR2 latent to split."}],"temporal_overlap":["INT",{"tooltip":"Latent frames shared between adjacent chunks and crossfaded at merge; 0 = no overlap.","default":0,"min":0,"max":16384}],"chunking_mode":["COMFY_DYNAMICCOMBO_V3",{"tooltip":"manual = use frames_per_chunk exactly; auto = predict the largest chunk that fits free VRAM.","options":[{"key":"auto","inputs":{"required":{}}},{"key":"manual","inputs":{"required":{"frames_per_chunk":["INT",{"tooltip":"Pixel frames per temporal chunk (4n+1: 1, 5, 9, 13, ...).","default":21,"min":1,"max":16384,"step":4}]}}}]}]}},"input_order":{"required":["latent","temporal_overlap","chunking_mode"]},"is_input_list":false,"output":["LATENT","INT"],"output_is_list":[true,false],"output_name":["latents","temporal_overlap"],"output_tooltips":["The temporal chunks in sequence order.","The effective latent-frame overlap between adjacent chunks, for Merge SeedVR2 Latents."],"output_matchtypes":null,"name":"SeedVR2TemporalChunk","display_name":"Split SeedVR2 Latent","description":"Split a SeedVR2 video latent into overlapping temporal chunks small enough to sample one at a time within VRAM, wiring latents outputs to both Apply SeedVR2 Conditioning and the sampler latent input before recombining with Merge SeedVR2 Latents.","python_module":"comfy_extras.nodes_seedvr","category":"model/latent/seedvr","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","split","chunk","temporal","video upscale","rebatch"],"essentials_category":null,"has_intermediate_output":false},"SeedVR2TemporalMerge":{"input":{"required":{"latents":["LATENT",{"tooltip":"The sampled temporal chunks in sequence order."}],"temporal_overlap":["INT",{"tooltip":"The temporal_overlap output of Split SeedVR2 Latent. 0 = plain concatenation.","default":0,"forceInput":true,"min":0,"max":16384}]}},"input_order":{"required":["latents","temporal_overlap"]},"is_input_list":true,"output":["LATENT"],"output_is_list":[false],"output_name":["latent"],"output_tooltips":["The recombined full-length latent."],"output_matchtypes":null,"name":"SeedVR2TemporalMerge","display_name":"Merge SeedVR2 Latents","description":"Recombine sampled SeedVR2 latent temporal chunks into one latent, crossfading each overlap with a Hann window sized by the temporal_overlap wired from Split SeedVR2 Latent.","python_module":"comfy_extras.nodes_seedvr","category":"model/latent/seedvr","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["seedvr2","merge","temporal","hann","crossfade"],"essentials_category":null,"has_intermediate_output":false},"SetLatentNoiseMask":{"input":{"required":{"samples":["LATENT"],"mask":["MASK"]}},"input_order":{"required":["samples","mask"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"SetLatentNoiseMask","display_name":"Set Latent Noise Mask","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"SolidMask":{"input":{"required":{"value":["FLOAT",{"default":1.0,"min":0.0,"max":1.0,"step":0.01}],"width":["INT",{"default":512,"min":1,"max":16384,"step":1}],"height":["INT",{"default":512,"min":1,"max":16384,"step":1}]}},"input_order":{"required":["value","width","height"]},"is_input_list":false,"output":["MASK"],"output_is_list":[false],"output_name":["MASK"],"output_tooltips":[null],"output_matchtypes":null,"name":"SolidMask","display_name":"Create Solid Mask","description":"","python_module":"comfy_extras.nodes_mask","category":"image/mask","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"StringConcatenate":{"input":{"required":{"string_a":["STRING",{"multiline":true}],"string_b":["STRING",{"multiline":true}],"delimiter":["STRING",{"default":"","multiline":false}]}},"input_order":{"required":["string_a","string_b","delimiter"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"StringConcatenate","display_name":"Concatenate Text","description":"","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["concatenate","text concat","join text","merge text","combine strings","string concat","append text","combine text"],"essentials_category":null,"has_intermediate_output":false},"StringContains":{"input":{"required":{"string":["STRING",{"multiline":true}],"substring":["STRING",{"multiline":true}],"case_sensitive":["BOOLEAN",{"advanced":true,"default":true}]}},"input_order":{"required":["string","substring","case_sensitive"]},"is_input_list":false,"output":["BOOLEAN"],"output_is_list":[false],"output_name":["contains"],"output_tooltips":[null],"output_matchtypes":null,"name":"StringContains","display_name":"Contains Text","description":"","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["contains","text includes","string includes"],"essentials_category":null,"has_intermediate_output":false},"StringReplace":{"input":{"required":{"string":["STRING",{"multiline":true}],"find":["STRING",{"multiline":true}],"replace":["STRING",{"multiline":true}]}},"input_order":{"required":["string","find","replace"]},"is_input_list":false,"output":["STRING"],"output_is_list":[false],"output_name":["STRING"],"output_tooltips":[null],"output_matchtypes":null,"name":"StringReplace","display_name":"Replace Text","description":"","python_module":"comfy_extras.nodes_string","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["replace","find and replace","substitute","swap text"],"essentials_category":null,"has_intermediate_output":false},"TextEncodeQwenImageEditPlus":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"multiline":true,"dynamicPrompts":true}]},"optional":{"vae":["VAE",{}],"image1":["IMAGE",{}],"image2":["IMAGE",{}],"image3":["IMAGE",{}]}},"input_order":{"required":["clip","prompt"],"optional":["vae","image1","image2","image3"]},"is_input_list":false,"output":["CONDITIONING"],"output_is_list":[false],"output_name":["CONDITIONING"],"output_tooltips":[null],"output_matchtypes":null,"name":"TextEncodeQwenImageEditPlus","display_name":"Text Encode Qwen Image Edit Plus","description":"","python_module":"comfy_extras.nodes_qwen","category":"model/conditioning/qwen image","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"TextGenerate":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"default":"","multiline":true,"dynamicPrompts":true}],"max_length":["INT",{"default":512,"min":1,"max":32768}],"sampling_mode":["COMFY_DYNAMICCOMBO_V3",{"display_name":"Sampling Mode","options":[{"key":"on","inputs":{"required":{"temperature":["FLOAT",{"default":0.7,"min":0.01,"max":2.0,"step":1e-06}],"top_k":["INT",{"default":64,"min":0,"max":1000}],"top_p":["FLOAT",{"default":0.95,"min":0.0,"max":1.0,"step":0.01}],"min_p":["FLOAT",{"default":0.05,"min":0.0,"max":1.0,"step":0.01}],"repetition_penalty":["FLOAT",{"default":1.05,"min":0.0,"max":5.0,"step":0.01}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615}]},"optional":{"presence_penalty":["FLOAT",{"default":0.0,"min":0.0,"max":5.0,"step":0.01}]}}},{"key":"off","inputs":{"required":{}}}]}]},"optional":{"image":["IMAGE",{}],"video":["IMAGE",{"tooltip":"Video frames as image batch. Assumed to be 24 FPS; subsampled to 1 FPS internally."}],"audio":["AUDIO",{}],"thinking":["BOOLEAN",{"tooltip":"Operate in thinking mode if the model supports it.","default":false}],"use_default_template":["BOOLEAN",{"tooltip":"Use the built in system prompt/template if the model has one.","advanced":true,"default":true}],"mtp":["COMBO",{"tooltip":"Speculative decoding with the checkpoint's multi-token-prediction head. No effect without MTP weights. auto adapts the draft depth; 2-5 pins it. Sampled output stays correctly distributed but differs from non-MTP output for the same seed.","default":"auto","multiselect":false,"options":["auto","off","2","3","4","5"]}],"system_prompt":["STRING",{"tooltip":"Replaces the system prompt in the model's chat template. Ignored when the default template is not used.","forceInput":true,"multiline":false}]}},"input_order":{"required":["clip","prompt","max_length","sampling_mode"],"optional":["image","video","audio","thinking","use_default_template","mtp","system_prompt"]},"is_input_list":false,"output":["STRING","STRING"],"output_is_list":[false,false],"output_name":["generated_text","thinking"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"TextGenerate","display_name":"Generate Text","description":"","python_module":"comfy_extras.nodes_textgen","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["LLM","gemma"],"essentials_category":null,"has_intermediate_output":false},"TextGenerateLTX2Prompt":{"input":{"required":{"clip":["CLIP",{}],"prompt":["STRING",{"default":"","multiline":true,"dynamicPrompts":true}],"max_length":["INT",{"default":512,"min":1,"max":32768}],"sampling_mode":["COMFY_DYNAMICCOMBO_V3",{"display_name":"Sampling Mode","options":[{"key":"on","inputs":{"required":{"temperature":["FLOAT",{"default":0.7,"min":0.01,"max":2.0,"step":1e-06}],"top_k":["INT",{"default":64,"min":0,"max":1000}],"top_p":["FLOAT",{"default":0.95,"min":0.0,"max":1.0,"step":0.01}],"min_p":["FLOAT",{"default":0.05,"min":0.0,"max":1.0,"step":0.01}],"repetition_penalty":["FLOAT",{"default":1.05,"min":0.0,"max":5.0,"step":0.01}],"seed":["INT",{"default":0,"min":0,"max":18446744073709551615}]},"optional":{"presence_penalty":["FLOAT",{"default":0.0,"min":0.0,"max":5.0,"step":0.01}]}}},{"key":"off","inputs":{"required":{}}}]}]},"optional":{"image":["IMAGE",{}],"video":["IMAGE",{"tooltip":"Video frames as image batch. Assumed to be 24 FPS; subsampled to 1 FPS internally."}],"audio":["AUDIO",{}],"thinking":["BOOLEAN",{"tooltip":"Operate in thinking mode if the model supports it.","default":false}],"use_default_template":["BOOLEAN",{"tooltip":"Use the built in system prompt/template if the model has one.","advanced":true,"default":true}],"mtp":["COMBO",{"tooltip":"Speculative decoding with the checkpoint's multi-token-prediction head. No effect without MTP weights. auto adapts the draft depth; 2-5 pins it. Sampled output stays correctly distributed but differs from non-MTP output for the same seed.","default":"auto","multiselect":false,"options":["auto","off","2","3","4","5"]}],"system_prompt":["STRING",{"tooltip":"Replaces the system prompt in the model's chat template. Ignored when the default template is not used.","forceInput":true,"multiline":false}]}},"input_order":{"required":["clip","prompt","max_length","sampling_mode"],"optional":["image","video","audio","thinking","use_default_template","mtp","system_prompt"]},"is_input_list":false,"output":["STRING","STRING"],"output_is_list":[false,false],"output_name":["generated_text","thinking"],"output_tooltips":[null,null],"output_matchtypes":null,"name":"TextGenerateLTX2Prompt","display_name":"Generate LTX2 Prompt","description":"","python_module":"comfy_extras.nodes_textgen","category":"text","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["prompt enhance","LLM","gemma"],"essentials_category":null,"has_intermediate_output":false},"TrimAudioDuration":{"input":{"required":{"audio":["AUDIO",{}],"start_index":["FLOAT",{"tooltip":"Start time in seconds, can be negative to count from the end (supports sub-seconds).","default":0.0,"min":-18446744073709551615,"max":18446744073709551615,"step":0.01}],"duration":["FLOAT",{"tooltip":"Duration in seconds","default":60.0,"min":0.0,"step":0.01}]}},"input_order":{"required":["audio","start_index","duration"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"TrimAudioDuration","display_name":"Trim Audio Duration","description":"Trim audio tensor into chosen time range.","python_module":"comfy_extras.nodes_audio","category":"audio","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["cut audio","audio clip","shorten audio"],"essentials_category":null,"has_intermediate_output":false},"TrimVideoLatent":{"input":{"required":{"samples":["LATENT",{}],"trim_amount":["INT",{"default":0,"min":0,"max":99999}]}},"input_order":{"required":["samples","trim_amount"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"TrimVideoLatent","display_name":"Trim Video Latent","description":"","python_module":"comfy_extras.nodes_wan","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"UNETLoader":{"input":{"required":{"unet_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors"]],"weight_dtype":[["default","fp8_e4m3fn","fp8_e4m3fn_fast","fp8_e5m2"],{"advanced":true}]}},"input_order":{"required":["unet_name","weight_dtype"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"name":"UNETLoader","display_name":"Load Diffusion Model","description":"","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"VAEDecode":{"input":{"required":{"samples":["LATENT",{"tooltip":"The latent to be decoded."}],"vae":["VAE",{"tooltip":"The VAE model used for decoding the latent."}]}},"input_order":{"required":["samples","vae"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"VAEDecode","display_name":"VAE Decode","description":"Decodes latent images back into pixel space images.","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"output_tooltips":["The decoded image."],"search_aliases":["decode","decode latent","latent to image","render latent"]},"VAEDecodeAudio":{"input":{"required":{"samples":["LATENT",{}],"vae":["VAE",{}]}},"input_order":{"required":["samples","vae"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"VAEDecodeAudio","display_name":"VAE Decode Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["latent to audio"],"essentials_category":null,"has_intermediate_output":false},"VAEDecodeAudioTiled":{"input":{"required":{"samples":["LATENT",{}],"vae":["VAE",{}],"tile_size":["INT",{"default":512,"min":32,"max":8192,"step":8}],"overlap":["INT",{"default":64,"min":0,"max":1024,"step":8}]}},"input_order":{"required":["samples","vae","tile_size","overlap"]},"is_input_list":false,"output":["AUDIO"],"output_is_list":[false],"output_name":["AUDIO"],"output_tooltips":[null],"output_matchtypes":null,"name":"VAEDecodeAudioTiled","display_name":"VAE Decode Audio (Tiled)","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["latent to audio"],"essentials_category":null,"has_intermediate_output":false},"VAEDecodeTiled":{"input":{"required":{"samples":["LATENT"],"vae":["VAE"],"tile_size":["INT",{"default":512,"min":64,"max":4096,"step":32,"advanced":true}],"overlap":["INT",{"default":64,"min":0,"max":4096,"step":32,"advanced":true}],"temporal_size":["INT",{"default":64,"min":8,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to decode at a time.","advanced":true}],"temporal_overlap":["INT",{"default":8,"min":4,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to overlap.","advanced":true}]}},"input_order":{"required":["samples","vae","tile_size","overlap","temporal_size","temporal_overlap"]},"is_input_list":false,"output":["IMAGE"],"output_is_list":[false],"output_name":["IMAGE"],"name":"VAEDecodeTiled","display_name":"VAE Decode (Tiled)","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"VAEEncode":{"input":{"required":{"pixels":["IMAGE"],"vae":["VAE"]}},"input_order":{"required":["pixels","vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"VAEEncode","display_name":"VAE Encode","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":["encode","encode image","image to latent"]},"VAEEncodeAudio":{"input":{"required":{"audio":["AUDIO",{}],"vae":["VAE",{}]}},"input_order":{"required":["audio","vae"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"output_tooltips":[null],"output_matchtypes":null,"name":"VAEEncodeAudio","display_name":"VAE Encode Audio","description":"","python_module":"comfy_extras.nodes_audio","category":"model/latent","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["audio to latent"],"essentials_category":null,"has_intermediate_output":false},"VAEEncodeTiled":{"input":{"required":{"pixels":["IMAGE"],"vae":["VAE"],"tile_size":["INT",{"default":512,"min":64,"max":4096,"step":64,"advanced":true}],"overlap":["INT",{"default":64,"min":0,"max":4096,"step":32,"advanced":true}],"temporal_size":["INT",{"default":64,"min":8,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to encode at a time.","advanced":true}],"temporal_overlap":["INT",{"default":8,"min":4,"max":4096,"step":4,"tooltip":"Only used for video VAEs: Amount of frames to overlap.","advanced":true}]}},"input_order":{"required":["pixels","vae","tile_size","overlap","temporal_size","temporal_overlap"]},"is_input_list":false,"output":["LATENT"],"output_is_list":[false],"output_name":["LATENT"],"name":"VAEEncodeTiled","display_name":"VAE Encode (Tiled)","description":"","python_module":"nodes","category":"model/latent","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"VAELoader":{"input":{"required":{"vae_name":[["Wan2_1_VAE_bf16.safetensors","ae.safetensors","birefnet.safetensors","clip_vision_h.safetensors","flux2-vae.safetensors","flux3-image-klein-base-int8-convrot.safetensors","flux3-image-klein-int8-convrot.safetensors","gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors","gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","gemma4_e2b_it_int8_convrot.safetensors","gemma_3_12B_it_fp4_mixed.safetensors","ltx-2.3-22b-dev-fp8.safetensors","ltx-2.3-22b-distilled-fp8.safetensors","ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","ltx-2.3-spatial-upscaler-x2-1.1.safetensors","ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","ltx-2.5-22b-ic-lora-clean-plate-1.0.safetensors","ltx-2.5-22b-ic-lora-day-to-night-0.9.safetensors","ltx-2.5-22b-ic-lora-water-simulation-0.9.safetensors","ltx-2.5-audio-vae-bf16.safetensors","ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","ltx-2.5-video-vae-bf16.safetensors","ltx2.3-transition.safetensors","ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors","minimax_music3_dav.safetensors","minimax_music3_dit_fp16.safetensors","minimax_music3_text_encoder_pruned_int8_convrot.safetensors","moge_2_vitl_normal_fp16.safetensors","qwen3.5_2b_bf16.safetensors","qwen3.5_9b_flux3_prompt_upsampler_w4a8.safetensors","qwen3vl_4b_int8_convrot.safetensors","qwen_3_4b_fp8_mixed.safetensors","sam3.1_multiplex_fp16.safetensors","seedvr2_3b_int8_convrot.safetensors","seedvr2_ema_vae_fp16.safetensors","stable_audio_3_medium.safetensors","t5gemma_b_b_ul2.safetensors","umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan_animate_2_distill_int8_convrot.safetensors","z_image_turbo_int8_convrot.safetensors","pixel_space"]]}},"input_order":{"required":["vae_name"]},"is_input_list":false,"output":["VAE"],"output_is_list":[false],"output_name":["VAE"],"name":"VAELoader","display_name":"Load VAE","description":"","python_module":"nodes","category":"model/loaders","output_node":false,"has_intermediate_output":false,"search_aliases":[]},"Video Slice":{"input":{"required":{"video":["VIDEO",{}],"start_time":["FLOAT",{"tooltip":"Start time in seconds","default":0.0,"min":-100000.0,"max":100000.0,"step":0.001}],"duration":["FLOAT",{"tooltip":"Duration in seconds, or 0 for unlimited duration","default":0.0,"min":0.0,"step":0.001}],"strict_duration":["BOOLEAN",{"tooltip":"If True, when the specified duration is not possible, an error will be raised.","default":false}]}},"input_order":{"required":["video","start_time","duration","strict_duration"]},"is_input_list":false,"output":["VIDEO"],"output_is_list":[false],"output_name":["VIDEO"],"output_tooltips":[null],"output_matchtypes":null,"name":"Video Slice","display_name":"Trim Video","description":"","python_module":"comfy_extras.nodes_video","category":"video","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":["trim video duration","skip first frames","frame load cap","start time"],"essentials_category":"Video Tools","has_intermediate_output":false},"WanAnimate2Cache":{"input":{"required":{"model":["MODEL",{}],"device":["COMBO",{"tooltip":"Where to keep the cache. cpu (RAM) is the safe choice, the cache will not fit in VRAM alongside the model at typical sizes. gpu (VRAM) can be faster if it fits.","default":"cpu","multiselect":false,"options":["cpu","gpu"]}],"dtype":["COMBO",{"tooltip":"Storage precision. default stores the activations in the model's compute dtype. int8 halves the cache, int4 quarters it, convrot is used to retain accuracy.","default":"default","multiselect":false,"options":["default","int8","int4"]}]}},"input_order":{"required":["model","device","dtype"]},"is_input_list":false,"output":["MODEL"],"output_is_list":[false],"output_name":["MODEL"],"output_tooltips":[null],"output_matchtypes":null,"name":"WanAnimate2Cache","display_name":null,"description":"Caches the pose-video's per-block activations so it runs once instead of on every sampling step. Roughly halves generation time Tradeoff is ~12.5 GB of system RAM at 480x832/81 frames in bf16 (scales with resolution and length). With context windows each window is cached separately, so RAM scales with the window count; use the static_standard schedule, as uniform schedules shift the windows every step and nothing ever recurs to hit the cache.","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan/animate","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanAnimate2ToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"tooltip":"Output video width in pixels.","default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"tooltip":"Output video height in pixels.","default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"tooltip":"Number of frames to generate.","default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"tooltip":"Number of videos to generate simultaneously.","default":1,"min":1,"max":4096}],"video_frame_offset":["INT",{"tooltip":"Frames to seek into the pose video. Connect to the video_frame_offset output of the previous node when extending.","default":0,"min":0,"max":16384,"step":1}],"pose_strength":["FLOAT",{"tooltip":"Scales the pose video's influence on the motion. 1.0 is the trained behavior; below weakens adherence, above amplifies. 0.0 mutes it but does not fully remove it.","default":1.0,"min":0.0,"max":10.0,"step":0.01}],"pose_start_percent":["FLOAT",{"tooltip":"Sampling percent at which the pose influence starts. Outside the window the pose branch is skipped entirely, which also speeds those steps up.","default":0.0,"min":0.0,"max":1.0,"step":0.01}],"pose_end_percent":["FLOAT",{"tooltip":"Sampling percent at which the pose influence ends. Motion is mostly established early, so e.g. 0.7 can loosen fine detail while keeping the choreography.","default":1.0,"min":0.0,"max":1.0,"step":0.01}],"reference_image_strength":["FLOAT",{"tooltip":"Scales how strongly generated frames attend to the reference image's latent frame. Below 1.0 loosens identity/appearance adherence (e.g. to let the prompt restyle), above tightens it against drift.","default":1.0,"min":0.0,"max":10.0,"step":0.01}]},"optional":{"reference_image":["IMAGE",{"tooltip":"The character to animate."}],"pose_video":["IMAGE",{"tooltip":"The video whose motion is transferred to the reference character."}],"clip_vision_output":["CLIP_VISION_OUTPUT",{"tooltip":"CLIP vision of the reference image."}],"positive_pose":["CONDITIONING",{"tooltip":"Prompt for the pose-video branch, describing the motion rather than the character. Defaults to positive. Used for both the cond and uncond passes."}],"clip_vision_output_pose":["CLIP_VISION_OUTPUT",{"tooltip":"CLIP vision of the pose video's first frame. Defaults to clip_vision_output."}],"continue_motion":["IMAGE",{"tooltip":"Previous motion sequence to continue from for temporal consistency."}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size","video_frame_offset","pose_strength","pose_start_percent","pose_end_percent","reference_image_strength"],"optional":["reference_image","pose_video","clip_vision_output","positive_pose","clip_vision_output_pose","continue_motion"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT","INT","INT","INT"],"output_is_list":[false,false,false,false,false,false],"output_name":["positive","negative","latent","trim_latent","trim_image","video_frame_offset"],"output_tooltips":[null,null,null,"Number of latent frames that should be trimmed before decoding.","Number of overlapping image frames when extending a video.","Frames to seek into the pose video."],"output_matchtypes":null,"name":"WanAnimate2ToVideo","display_name":null,"description":"Animate a character in a reference image using a video, effectively replicating the facial animation, body motion and hands gesture from the video.","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan/animate","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanCameraEmbedding":{"input":{"required":{"camera_pose":["COMBO",{"default":"Static","multiselect":false,"options":["Static","Pan Up","Pan Down","Pan Left","Pan Right","Zoom In","Zoom Out","Anti Clockwise (ACW)","ClockWise (CW)"]}],"width":["INT",{"default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}]},"optional":{"speed":["FLOAT",{"default":1.0,"min":0,"max":10.0,"step":0.1}],"fx":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":1e-09}],"fy":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":1e-09}],"cx":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":0.01}],"cy":["FLOAT",{"advanced":true,"default":0.5,"min":0,"max":1,"step":0.01}]}},"input_order":{"required":["camera_pose","width","height","length"],"optional":["speed","fx","fy","cx","cy"]},"is_input_list":false,"output":["WAN_CAMERA_EMBEDDING","INT","INT","INT"],"output_is_list":[false,false,false,false],"output_name":["camera_embedding","width","height","length"],"output_tooltips":[null,null,null,null],"output_matchtypes":null,"name":"WanCameraEmbedding","display_name":null,"description":"","python_module":"comfy_extras.nodes_camera_trajectory","category":"model/conditioning/wan/camera","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanCameraImageToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]},"optional":{"clip_vision_output":["CLIP_VISION_OUTPUT",{}],"start_image":["IMAGE",{}],"camera_conditions":["WAN_CAMERA_EMBEDDING",{}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size"],"optional":["clip_vision_output","start_image","camera_conditions"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"WanCameraImageToVideo","display_name":null,"description":"","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan/camera","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanImageToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"default":832,"min":16,"max":16384,"step":16}],"height":["INT",{"default":480,"min":16,"max":16384,"step":16}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"default":1,"min":1,"max":4096}]},"optional":{"clip_vision_output":["CLIP_VISION_OUTPUT",{}],"start_image":["IMAGE",{}],"ref_pad_image":["IMAGE",{"tooltip":"Fills the padding frames of the image conditioning with this image instead of gray, anchoring identity without pinning frames (SVI-style anti-drift padding, used by models such as ID-V2V)."}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size"],"optional":["clip_vision_output","start_image","ref_pad_image"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT"],"output_is_list":[false,false,false],"output_name":["positive","negative","latent"],"output_tooltips":[null,null,null],"output_matchtypes":null,"name":"WanImageToVideo","display_name":null,"description":"","python_module":"comfy_extras.nodes_wan","category":"model/conditioning/wan","output_node":false,"deprecated":false,"experimental":false,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false},"WanSCAILToVideo":{"input":{"required":{"positive":["CONDITIONING",{}],"negative":["CONDITIONING",{}],"vae":["VAE",{}],"width":["INT",{"default":512,"min":32,"max":16384,"step":32}],"height":["INT",{"default":896,"min":32,"max":16384,"step":32}],"length":["INT",{"default":81,"min":1,"max":16384,"step":4}],"batch_size":["INT",{"default":1,"min":1,"max":4096}],"pose_strength":["FLOAT",{"tooltip":"Strength of the pose latent.","default":1.0,"min":0.0,"max":10.0,"step":0.01}],"pose_start":["FLOAT",{"tooltip":"Start step of the pose conditioning.","default":0.0,"min":0.0,"max":1.0,"step":0.01}],"pose_end":["FLOAT",{"tooltip":"End step of the pose conditioning.","default":1.0,"min":0.0,"max":1.0,"step":0.01}],"video_frame_offset":["INT",{"tooltip":"Cumulative output frame this chunk begins at. Wire from the previous chunk's video_frame_offset output.","default":0,"min":0,"max":16384,"step":1}],"previous_frame_count":["INT",{"tooltip":"Tail frames of previous_frames to anchor. SCAIL-2 trained at 5 (81-frame chunks, 76-frame step).","default":5,"min":1,"max":16384,"step":4}]},"optional":{"pose_video":["IMAGE",{"tooltip":"Video used for pose conditioning. Will be downscaled to half the resolution of the main video."}],"pose_video_mask":["IMAGE",{"tooltip":"SCAIL-2 only. Colored per-identity SAM3 mask video at the same resolution as pose_video."}],"replacement_mode":["BOOLEAN",{"tooltip":"SCAIL-2 only. False = Animation Mode (pose_video_mask should have black background). True = Replacement Mode (pose_video_mask should have white background).","default":false}],"reference_image":["IMAGE",{"tooltip":"Reference image. The first image is the primary reference (composite all identities onto it). SCAIL-2: extra batch images are used as additional views (back view, close-up, occluded background), each needing a matching reference_image_mask in that identity's color."}],"reference_image_mask":["IMAGE",{"tooltip":"SCAIL-2 only. Colored reference mask, batch matching reference_image (first = primary reference mask, rest = identity masks for the additional reference_image)."}],"clip_vision_output":["CLIP_VISION_OUTPUT",{"tooltip":"CLIP vision features for conditioning. Model is trained with stretch resize to aspect ratio."}],"previous_frames":["IMAGE",{"tooltip":"SCAIL-2 only. Full decoded output of the previous chunk. Only the last previous_frame_count are used as the extension anchor."}]}},"input_order":{"required":["positive","negative","vae","width","height","length","batch_size","pose_strength","pose_start","pose_end","video_frame_offset","previous_frame_count"],"optional":["pose_video","pose_video_mask","replacement_mode","reference_image","reference_image_mask","clip_vision_output","previous_frames"]},"is_input_list":false,"output":["CONDITIONING","CONDITIONING","LATENT","INT"],"output_is_list":[false,false,false,false],"output_name":["positive","negative","latent","video_frame_offset"],"output_tooltips":[null,null,"Empty latent of the generation size.","Adjusted offset + length. Wire into the next chunk."],"output_matchtypes":null,"name":"WanSCAILToVideo","display_name":null,"description":"","python_module":"comfy_extras.nodes_scail","category":"model/conditioning/wan/scail","output_node":false,"deprecated":false,"experimental":true,"dev_only":false,"api_node":false,"price_badge":null,"search_aliases":null,"essentials_category":null,"has_intermediate_output":false}}
diff --git a/tools/template_apps.py b/tools/template_apps.py
index 889d68e..ebe83c5 100644
--- a/tools/template_apps.py
+++ b/tools/template_apps.py
@@ -18,6 +18,11 @@ these bases, all workflows built by tools/build_apps.py:
morph FLUX 3 Klein draws the last frame from the photo and LTX-2.5 + the transition LoRA moves
the photo into it (apps/morph): for clips where the scene must change around a subject
that has to stay the same (a face, a product); "klein" is that fixed instruction
+ effect Wan 2.1 image-to-video with one effect LoRA fixed (apps/effects); "effect_lora" names it
+ effect-klein the same, after FLUX 3 Klein redraws the photo with the fixed "klein" instruction
+ (merged in as edit:*): for effects that need another background or something added first
+ edit + outpaint the edit pipeline on a padded canvas: the sampler starts from the photo with gray borders
+ and a noise mask on them, so only the new border is drawn; a select sets the padding
The app folder gets the base workflow (copied, or with the few edits noted in BASES) and a manifest
whose "details" input carries the prompt as a bind template; selects feed {placeholders} in that
prompt. The base app's models and credits are copied as they are (qwen-edit swaps the LoRA).
@@ -50,7 +55,14 @@ BASES = {
"output": {"node": "12", "type": "image"}},
"morph": {"app": "morph", "section": "video", "pack": "video-ltx25", "image": ("31", "image"), "prompt": ("251:252", "value"), "klein": ("edit:4", "text"),
"keep": ("aspect", "duration", "strength", "seed"), "output": {"node": "68", "type": "video"}},
+ "effect": {"app": "effects", "section": "video", "pack": "video-wan-effects", "image": ("52", "image"), "prompt": ("6", "text"),
+ "keep": ("strength", "aspect", "frames", "seed"), "output": {"node": "56", "type": "video"}},
+ "effect-klein": {"app": "effects", "section": "video", "pack": "video-wan-effects", "image": ("52", "image"), "prompt": ("6", "text"),
+ "klein": ("edit:4", "text"), "keep": ("strength", "aspect", "frames", "seed"), "output": {"node": "56", "type": "video"}},
}
+BULLET_TIME = "Wan21_I2V_rotation_000003750.safetensors" # joyfox's Bullet Time LoRA (Apache-2.0), the effects app's "Bullet time"
+BULLET_TRIGGER = "rw360xz,人物定在空中,相机视角围绕人物向一个方向旋转。" # its trigger words, as on its model card
+
# MediaPipe Face Landmarker as ComfyUI's native nodes load it (comfy_extras/nodes_mediapipe.py, blueprint "Image Face Detection")
FACE_MODEL = {"repo": "Comfy-Org/mediapipe", "file": "detection/mediapipe_face_fp32.safetensors", "dir": "detection", "gb": 0.01, "phase": "encode",
@@ -104,12 +116,18 @@ T = []
def A(id_, name, feature, cat, base, image_label, summary, prompt, inputs=(), ref=None, values=None, note="", image2_label=None, ref2=None, lora=None, help_=None,
- defaults=None, klein=None, face_mask=False):
- """defaults: new defaults for settings kept from the base app (say, the LTX enhancer off); klein: the morph base's Klein instruction;
- face_mask (edit2): redraw only the face oval MediaPipe finds in the first image, so the rest of it stays as it is."""
+ defaults=None, klein=None, face_mask=False, effect_lora=None, outpaint=False):
+ """defaults: new defaults for settings kept from the base app (say, the LTX enhancer off); klein: the morph and effect-klein bases' Klein
+ instruction; face_mask (edit2): redraw only the face oval MediaPipe finds in the first image, so the rest of it stays as it is;
+ effect_lora: the effect bases' LoRA file; outpaint (edit): draw only a border added around the photo (a select's sets pad it)."""
T.append(dict(id=id_, name=name, feature=feature, cat=cat, base=base, image_label=image_label, summary=summary, prompt=prompt,
inputs=list(inputs), ref=ref, values=values or {}, note=note, image2_label=image2_label, ref2=ref2, lora=lora, help=help_,
- defaults=defaults or {}, klein=klein, face_mask=face_mask))
+ defaults=defaults or {}, klein=klein, face_mask=face_mask, effect_lora=effect_lora, outpaint=outpaint))
+
+
+def pad(left, top, right, bottom):
+ """The sets of an outpaint select option: pixels added on each side of the photo (scaled to 1 MP first)."""
+ return [{"node": "op_pad", "field": f, "value": v} for f, v in (("left", left), ("top", top), ("right", right), ("bottom", bottom))]
# ---------------------------------------------------------------- ads & products (apps/edit pipeline, a product photo)
@@ -483,6 +501,53 @@ A("idol", "Idol", "Idol", "clips", "video", "Your photo",
"crowd cheering and a pop beat.",
[], WOMAN, {"aspect": PORTRAIT}, "Stage, confetti, pose.")
+# ---------------------------------------------------------------- the canvas, a new person, one effect LoRA
+A("expand-image", "Expand Image", "Expand Image", "style", "edit", "Your photo",
+ "Grows a photo past its edges: FLUX 3 Image Klein draws only the new border, so the original stays as it was.",
+ "Extend this photo outward into the gray border around it: continue the scene, its objects, light and perspective naturally past the "
+ "original edges, as if the camera had pulled back. The original part of the photo stays exactly the same; no frame, no border, no text.",
+ [{"id": "direction", "type": "select", "label": "Expand", "default": "out", "options": [
+ {"value": "out", "label": "Zoom out", "sets": pad(192, 192, 192, 192)},
+ {"value": "wide", "label": "Wider", "sets": pad(320, 0, 320, 0)},
+ {"value": "tall", "label": "Taller", "sets": pad(0, 320, 0, 320)}]}],
+ LAKE, {"direction": "out"}, "Does the border continue the lake, rock and sky, with no seam or gray left?", outpaint=True)
+A("ai-influencer", "AI Influencer", "Influencer IA", "portraits", "z-image", None,
+ "Makes a feed-ready photo of a made-up influencer in their niche, with natural skin (Z-Image Turbo).",
+ "A vertical phone photo for a social media feed of {who}, a {niche}. Natural skin texture with visible pores, candid moment, "
+ "realistic photo, no text, no logos.",
+ [txt("who", "Who", "a woman in her late twenties from Mexico City with curly dark hair", required=True,
+ help_="Describe a person who does not exist: age, origin, hair, style."),
+ sel("niche", "Niche", [("Fitness", "fitness influencer stretching on a rooftop at sunrise in workout clothes"),
+ ("Travel", "travel influencer with a backpack at a viewpoint above a coastal town"),
+ ("Fashion", "fashion influencer posing on a city street in a tailored camel coat"),
+ ("Food", "food influencer tasting a dish at a busy street food stall at night"),
+ ("Tech", "tech influencer at a desk with a laptop and a camera on a tripod, warm desk lamp")])],
+ None, {"niche": "fitness influencer stretching on a rooftop at sunrise in workout clothes"}, "A believable person, not a glossy render?",
+ defaults={"width": 768, "height": 1280})
+A("bullet-time-scene", "Bullet Time Scene", "Bullet Time Scene", "camera", "effect", "Your photo",
+ "Freezes the moment and circles the camera around the subject, in their own scene (Wan 2.1 + Bullet Time LoRA).",
+ BULLET_TRIGGER + "The video shows {subject}, frozen in mid-air while the camera circles around them in one direction; the whole scene is frozen too.",
+ [txt("subject", "Who or what is in the photo", "a person mid-jump", required=True)],
+ "ref-fx-jump", {"subject": "a parkour athlete in a black t-shirt leaping between two rooftops"},
+ "Does the camera orbit while the athlete and the city stay frozen?", effect_lora=BULLET_TIME)
+A("bullet-time-white", "Bullet Time White", "Bullet Time White", "camera", "effect-klein", "Your photo",
+ "Puts the subject mid-jump on a white studio backdrop, then freezes time and circles the camera around them.",
+ BULLET_TRIGGER + "The video shows {subject}, frozen in mid-air on a seamless white studio background while the camera circles around them in one direction.",
+ [txt("subject", "Who or what is in the photo", "a person mid-jump", required=True)],
+ CHARACTER, {"subject": "a man in a red jacket, jeans and white sneakers"},
+ "White backdrop, same person and outfit, smooth orbit?", effect_lora=BULLET_TIME,
+ klein="Put this exact person on a seamless pure white studio background, mid-jump with arms spread, soft even studio light. "
+ "Keep the face, hair and outfit identical.")
+A("bullet-time-splash", "Bullet Time Splash", "Bullet Time Splash", "camera", "effect-klein", "Your photo",
+ "Bursts frozen water around the subject, then circles the camera through the suspended drops.",
+ BULLET_TRIGGER + "The video shows {subject}, frozen in mid-air inside a burst of suspended water droplets while the camera circles around them in one direction.",
+ [txt("subject", "Who or what is in the photo", "a person", required=True)],
+ "ref-fx-runner", {"subject": "a sprinter in a red and white running outfit"},
+ "Do the drops stay frozen in the air while the camera moves?", effect_lora=BULLET_TIME,
+ klein="Add a frozen splash of clear water bursting around this exact person, droplets and ribbons of water suspended in the air, "
+ "same place and light. Keep the face, hair and outfit identical.")
+
+
# ---------------------------------------------------------------- camera LoRAs on Qwen-Image-Edit-2509 (the relight pipeline, other LoRA)
MOVES = [("Move forward", "将镜头向前移动 Move the camera forward."), ("Move back", "将镜头向后移动 Move the camera backward."),
("Move left", "将镜头向左移动 Move the camera left."), ("Move right", "将镜头向右移动 Move the camera right."),
@@ -534,8 +599,9 @@ def check() -> None:
assert t["base"] in BASES, t["id"]
used = set(re.findall(r"\{([a-z][a-z0-9_]*)\}", t["prompt"] + " " + (t["klein"] or "")))
have = {i["id"] for i in t["inputs"]}
+ setters = {i["id"] for i in t["inputs"] if any("sets" in o for o in i.get("options", []))} # selects that write node fields
assert used <= have, (t["id"], used - have)
- assert have <= used, (t["id"], have - used)
+ assert have - setters <= used, (t["id"], have - setters - used)
for i in t["inputs"]:
assert i["id"] not in ("image", "image2", "details", "steps", "seed", "width", "height", "duration", "enhance", "aspect"), (t["id"], i["id"])
if t["base"] == "z-image":
@@ -544,7 +610,9 @@ def check() -> None:
assert t["ref"], t["id"]
assert (t["base"] == "edit2") == bool(t["image2_label"] and t["ref2"]), t["id"]
assert (t["base"] == "qwen-edit") == (t["lora"] in QWEN_LORAS), t["id"]
- assert (t["base"] == "morph") == bool(t["klein"]), t["id"]
+ assert (t["base"] in ("morph", "effect-klein")) == bool(t["klein"]), t["id"]
+ assert (t["base"] in ("effect", "effect-klein")) == bool(t["effect_lora"]), t["id"]
+ assert not t["outpaint"] or t["base"] == "edit", t["id"]
assert not t["face_mask"] or t["base"] == "edit2", t["id"]
for k in t["defaults"]:
assert k in BASES[t["base"]]["keep"], (t["id"], k)
@@ -588,6 +656,19 @@ def manifest(t: dict) -> dict:
credits = credits + ["MediaPipe Face Landmarker (Google, Apache-2.0), ComfyUI port: Comfy-Org/mediapipe"]
if t["base"] == "morph":
notes += f" with a fixed Klein instruction for the last frame (node edit:4: \"{t['klein']}\")"
+ if t["base"] in ("effect", "effect-klein"):
+ effects = {s["value"] for i in src["inputs"] if i["id"] == "effect" for o in i["options"] for s in o.get("sets", []) if s["field"] == "lora_name"}
+ models = [m for m in models if m["file"].split("/")[-1] not in effects or m["file"].split("/")[-1] == t["effect_lora"]]
+ notes += f" with {t['effect_lora']} as the effect LoRA (node ff_effect)"
+ if t["base"] == "effect-klein":
+ edit = json.loads((ROOT / "apps" / "edit" / "app.json").read_text(encoding="utf-8"))
+ models = models + edit["models"]
+ credits = credits + [c for c in edit.get("credits", []) if c not in credits]
+ notes += (f", after FLUX 3 Klein (apps/edit, merged in as edit:*) redraws the photo with a fixed instruction (node edit:4: \"{t['klein']}\"); "
+ "Wan starts from Klein's image")
+ if t["outpaint"]:
+ notes += ("; the sampler starts from the photo scaled to 1 MP and padded by ImagePadForOutpaint (the select sets the padding), with a noise "
+ "mask on the border only, and the unpadded photo stays Klein's reference")
if t["lora"]:
lora = QWEN_LORAS[t["lora"]]
models = [m for m in models if "Relight" not in m["file"]] + [{"repo": lora["repo"], "file": lora["file"], "dir": "loras", "gb": lora["gb"], "phase": "denoise", "license": lora["license"]}]
@@ -610,6 +691,10 @@ def manifest(t: dict) -> dict:
}
+def is_link(v) -> bool:
+ return isinstance(v, list) and len(v) == 2 and isinstance(v[0], str)
+
+
def workflow(t: dict) -> dict:
"""The base app's workflow, with the few edits a base needs."""
base = BASES[t["base"]]
@@ -635,7 +720,33 @@ def workflow(t: dict) -> dict:
api["7"]["inputs"]["latent_image"] = ["fm_latent", 0]
api["17"]["inputs"]["images.image0"] = ["12", 0] # the face photo is the only reference: with both, Klein copied the first face back
del api["17"]["inputs"]["images.image1"]
- if t["base"] == "morph": # the last frame Klein draws is fixed per template (placeholders filled with the defaults)
+ if t["base"] in ("effect", "effect-klein"):
+ api["ff_effect"]["inputs"]["lora_name"] = t["effect_lora"]
+ if t["base"] == "effect-klein": # Klein redraws the photo first; Wan's start image and CLIP vision read Klein's output
+ edit = json.loads((ROOT / "apps" / "edit" / "workflow.api.json").read_text(encoding="utf-8"))
+ for i, n in edit.items():
+ node = json.loads(json.dumps(n))
+ for k, v in node["inputs"].items():
+ if is_link(v):
+ node["inputs"][k] = [f"edit:{v[0]}", v[1]]
+ api[f"edit:{i}"] = node
+ del api["edit:9"], api["edit:11"] # Klein's own save, and its own LoadImage: it reads the app's photo
+ for i, n in api.items():
+ for k, v in n["inputs"].items():
+ if is_link(v) and v[0] == "edit:11":
+ n["inputs"][k] = ["52", 0]
+ elif is_link(v) and v[0] == "52" and not i.startswith("edit:"):
+ n["inputs"][k] = ["edit:8", 0]
+ if t["outpaint"]: # draw only the border: start from the padded photo, noise mask on the padding, the photo as reference
+ api["op_scale"] = {"class_type": "ImageScaleToTotalPixels", "inputs": {"image": ["11", 0], "upscale_method": "lanczos", "megapixels": 1.0,
+ "resolution_steps": 16}, "_meta": {"title": "Photo at 1 MP"}}
+ api["op_pad"] = {"class_type": "ImagePadForOutpaint", "inputs": {"image": ["op_scale", 0], "left": 192, "top": 192, "right": 192, "bottom": 192,
+ "feathering": 24}, "_meta": {"title": "Border to draw"}}
+ api["op_encode"] = {"class_type": "VAEEncode", "inputs": {"pixels": ["op_pad", 0], "vae": ["3", 0]}, "_meta": {"title": "Encode the padded photo"}}
+ api["op_latent"] = {"class_type": "SetLatentNoiseMask", "inputs": {"samples": ["op_encode", 0], "mask": ["op_pad", 1]},
+ "_meta": {"title": "Only the border is drawn"}}
+ api["7"]["inputs"]["latent_image"] = ["op_latent", 0]
+ if t["base"] in ("morph", "effect-klein"): # the image Klein draws is fixed per template (placeholders filled with the defaults)
defaults = {i["id"]: i.get("default", "") for i in t["inputs"]}
api[base["klein"][0]]["inputs"][base["klein"][1]] = re.sub(r"\{([a-z][a-z0-9_]*)\}", lambda m: str(defaults.get(m.group(1), "")), t["klein"])
if t["lora"]: # the relight pipeline with another LoRA; node 14 prepends the Relight trigger, so it is blanked
diff --git a/web/js/catalog.js b/web/js/catalog.js
index 6bdf479..0fdb27a 100644
--- a/web/js/catalog.js
+++ b/web/js/catalog.js
@@ -22,14 +22,14 @@ export async function loadCatalog() {
/** Apps grouped like a store: a card can only sit in one row. */
export const CATEGORIES = [
- { id: "create", apps: ["z-image", "cinema-studio", "styles", "video-t2v", "video", "music", "sfx"] },
+ { id: "create", apps: ["z-image", "cinema-studio", "styles", "video-t2v", "video", "h3-video", "h3-photo", "music", "sfx"] },
{ id: "identity", apps: ["character", "speak", "talking", "recast", "motion-transfer", "genjutsu", "dub", "character-swap", "face-swap"] },
- { id: "camera", apps: ["camera-moves", "camera-copy", "reshoot", "reframe", "transitions", "angles", "shots", "zooms", "whats-next"] },
- { id: "style", apps: ["relight", "style-reference", "mixed-media", "upscale", "upscale-video", "sdr-hdr", "smooth", "remove-bg"] },
+ { id: "camera", apps: ["camera-moves", "camera-copy", "reshoot", "reframe", "extend-video", "transitions", "angles", "shots", "zooms", "whats-next", "bullet-time-scene", "bullet-time-white", "bullet-time-splash"] },
+ { id: "style", apps: ["relight", "style-reference", "mixed-media", "upscale", "upscale-video", "sdr-hdr", "smooth", "remove-bg", "video-remove-bg", "depth-map", "expand-image"] },
{ id: "vfx", apps: ["effects", "morph", "edit", "vfx-clean-plate", "vfx-day-to-night", "vfx-water"] },
{ id: "ads", apps: ["product-shot", "poster", "packshot", "macro-scene", "macro-product", "giant-product", "billboard", "graffiti-ad", "fridge-ad", "ugc-shot",
"asmr-product", "chameleon", "magic-button", "truck-ad", "kick-ad", "volcano-ad"] },
- { id: "portraits", apps: ["headshot", "commercial-faces", "style-snap", "skin-enhancer", "outfit-vending", "glitter-sticker", "fashion-factory", "try-on", "outfit-swap", "outfit-shot"] },
+ { id: "portraits", apps: ["ai-influencer", "headshot", "commercial-faces", "style-snap", "skin-enhancer", "outfit-vending", "glitter-sticker", "fashion-factory", "try-on", "outfit-swap", "outfit-shot"] },
{ id: "clips", apps: ["behind-the-scenes", "asmr-whisper", "asmr-host", "roller-coaster", "storm-creature", "cloud-surf", "mukbang", "melting-doodle", "rap-video", "idol"] },
{ id: "play", apps: ["game-styles", "plushie", "render-3d", "figure-3d", "pixel-game", "brick-toy", "sketch-to-real", "paint-app", "mural", "micro-beasts", "meme"] },
{ id: "templates", apps: ["j-magazine", "cafe-60s", "victory-card", "j-poster", "mascot", "burning-sunset", "comic-book", "giallo", "renaissance", "japanese-show", "breakdown", "edit-text", "image-effects", "color-grade"] },