Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

api nodes(Ideogram): add Ideogram Character (#9616)

Alexander Piskun50333f1715c0
diff --git a/comfy_api_nodes/apis/__init__.py b/comfy_api_nodes/apis/__init__.py
index 7a09df55b..78a23db30 100644
--- a/comfy_api_nodes/apis/__init__.py
+++ b/comfy_api_nodes/apis/__init__.py
@@ -951,7 +951,11 @@ class MagicPrompt2(str, Enum):
class StyleType1(str, Enum):
+ AUTO = 'AUTO'
GENERAL = 'GENERAL'
+ REALISTIC = 'REALISTIC'
+ DESIGN = 'DESIGN'
+ FICTION = 'FICTION'
class ImagenImageGenerationInstance(BaseModel):
@@ -2676,7 +2680,7 @@ class ReleaseNote(BaseModel):
class RenderingSpeed(str, Enum):
- BALANCED = 'BALANCED'
+ DEFAULT = 'DEFAULT'
TURBO = 'TURBO'
QUALITY = 'QUALITY'
@@ -4918,6 +4922,14 @@ class IdeogramV3EditRequest(BaseModel):
None,
description='A set of images to use as style references (maximum total size 10MB across all style references). The images should be in JPEG, PNG or WebP format.',
)
+ character_reference_images: Optional[List[str]] = Field(
+ None,
+ description='Generations with character reference are subject to the character reference pricing. A set of images to use as character references (maximum total size 10MB across all character references), currently only supports 1 character reference image. The images should be in JPEG, PNG or WebP format.'
+ )
+ character_reference_images_mask: Optional[List[str]] = Field(
+ None,
+ description='Optional masks for character reference images. When provided, must match the number of character_reference_images. Each mask should be a grayscale image of the same dimensions as the corresponding character reference image. The images should be in JPEG, PNG or WebP format.'
+ )
class IdeogramV3Request(BaseModel):
@@ -4951,6 +4963,14 @@ class IdeogramV3Request(BaseModel):
style_type: Optional[StyleType1] = Field(
None, description='The type of style to apply'
)
+ character_reference_images: Optional[List[str]] = Field(
+ None,
+ description='Generations with character reference are subject to the character reference pricing. A set of images to use as character references (maximum total size 10MB across all character references), currently only supports 1 character reference image. The images should be in JPEG, PNG or WebP format.'
+ )
+ character_reference_images_mask: Optional[List[str]] = Field(
+ None,
+ description='Optional masks for character reference images. When provided, must match the number of character_reference_images. Each mask should be a grayscale image of the same dimensions as the corresponding character reference image. The images should be in JPEG, PNG or WebP format.'
+ )
class ImagenGenerateImageResponse(BaseModel):
diff --git a/comfy_api_nodes/nodes_ideogram.py b/comfy_api_nodes/nodes_ideogram.py
index d28895f3e..2d1c32e4f 100644
--- a/comfy_api_nodes/nodes_ideogram.py
+++ b/comfy_api_nodes/nodes_ideogram.py
@@ -255,6 +255,7 @@ class IdeogramV1(comfy_io.ComfyNode):
display_name="Ideogram V1",
category="api node/image/Ideogram",
description="Generates images using the Ideogram V1 model.",
+ is_api_node=True,
inputs=[
comfy_io.String.Input(
"prompt",
@@ -383,6 +384,7 @@ class IdeogramV2(comfy_io.ComfyNode):
display_name="Ideogram V2",
category="api node/image/Ideogram",
description="Generates images using the Ideogram V2 model.",
+ is_api_node=True,
inputs=[
comfy_io.String.Input(
"prompt",
@@ -552,6 +554,7 @@ class IdeogramV3(comfy_io.ComfyNode):
category="api node/image/Ideogram",
description="Generates images using the Ideogram V3 model. "
"Supports both regular image generation from text prompts and image editing with mask.",
+ is_api_node=True,
inputs=[
comfy_io.String.Input(
"prompt",
@@ -612,11 +615,21 @@ class IdeogramV3(comfy_io.ComfyNode):
),
comfy_io.Combo.Input(
"rendering_speed",
- options=["BALANCED", "TURBO", "QUALITY"],
- default="BALANCED",
+ options=["DEFAULT", "TURBO", "QUALITY"],
+ default="DEFAULT",
tooltip="Controls the trade-off between generation speed and quality",
optional=True,
),
+ comfy_io.Image.Input(
+ "character_image",
+ tooltip="Image to use as character reference.",
+ optional=True,
+ ),
+ comfy_io.Mask.Input(
+ "character_mask",
+ tooltip="Optional mask for character reference image.",
+ optional=True,
+ ),
],
outputs=[
comfy_io.Image.Output(),
@@ -639,12 +652,46 @@ class IdeogramV3(comfy_io.ComfyNode):
magic_prompt_option="AUTO",
seed=0,
num_images=1,
- rendering_speed="BALANCED",
+ rendering_speed="DEFAULT",
+ character_image=None,
+ character_mask=None,
):
auth = {
"auth_token": cls.hidden.auth_token_comfy_org,
"comfy_api_key": cls.hidden.api_key_comfy_org,
}
+ if rendering_speed == "BALANCED": # for backward compatibility
+ rendering_speed = "DEFAULT"
+
+ character_img_binary = None
+ character_mask_binary = None
+
+ if character_image is not None:
+ input_tensor = character_image.squeeze().cpu()
+ if character_mask is not None:
+ character_mask = resize_mask_to_image(character_mask, character_image, allow_gradient=False)
+ character_mask = 1.0 - character_mask
+ if character_mask.shape[1:] != character_image.shape[1:-1]:
+ raise Exception("Character mask and image must be the same size")
+
+ mask_np = (character_mask.squeeze().cpu().numpy() * 255).astype(np.uint8)
+ mask_img = Image.fromarray(mask_np)
+ mask_byte_arr = BytesIO()
+ mask_img.save(mask_byte_arr, format="PNG")
+ mask_byte_arr.seek(0)
+ character_mask_binary = mask_byte_arr
+ character_mask_binary.name = "mask.png"
+
+ img_np = (input_tensor.numpy() * 255).astype(np.uint8)
+ img = Image.fromarray(img_np)
+ img_byte_arr = BytesIO()
+ img.save(img_byte_arr, format="PNG")
+ img_byte_arr.seek(0)
+ character_img_binary = img_byte_arr
+ character_img_binary.name = "image.png"
+ elif character_mask is not None:
+ raise Exception("Character mask requires character image to be present")
+
# Check if both image and mask are provided for editing mode
if image is not None and mask is not None:
# Edit mode
@@ -693,6 +740,15 @@ class IdeogramV3(comfy_io.ComfyNode):
if num_images > 1:
edit_request.num_images = num_images
+ files = {
+ "image": img_binary,
+ "mask": mask_binary,
+ }
+ if character_img_binary:
+ files["character_reference_images"] = character_img_binary
+ if character_mask_binary:
+ files["character_mask_binary"] = character_mask_binary
+
# Execute the operation for edit mode
operation = SynchronousOperation(
endpoint=ApiEndpoint(
@@ -702,10 +758,7 @@ class IdeogramV3(comfy_io.ComfyNode):
response_model=IdeogramGenerateResponse,
),
request=edit_request,
- files={
- "image": img_binary,
- "mask": mask_binary,
- },
+ files=files,
content_type="multipart/form-data",
auth_kwargs=auth,
)
@@ -739,6 +792,14 @@ class IdeogramV3(comfy_io.ComfyNode):
if num_images > 1:
gen_request.num_images = num_images
+ files = {}
+ if character_img_binary:
+ files["character_reference_images"] = character_img_binary
+ if character_mask_binary:
+ files["character_mask_binary"] = character_mask_binary
+ if files:
+ gen_request.style_type = "AUTO"
+
# Execute the operation for generation mode
operation = SynchronousOperation(
endpoint=ApiEndpoint(
@@ -748,6 +809,8 @@ class IdeogramV3(comfy_io.ComfyNode):
response_model=IdeogramGenerateResponse,
),
request=gen_request,
+ files=files if files else None,
+ content_type="multipart/form-data",
auth_kwargs=auth,
)