Skip to content

World's first open-source, agentic video production system. 12 production pipelines, 100+ tools, 700+ agent skill and production-knowledge files. Turn your AI coding assistant into a full video production studio.

fix(comfyui): use WAN 2.1 VAE in bundled 14B T2V workflow

Mandaar37e4ef783172
diff --git a/tools/_comfyui/workflows/wan22-t2v-4step.json b/tools/_comfyui/workflows/wan22-t2v-4step.json
index f5772aca..5016de5e 100644
--- a/tools/_comfyui/workflows/wan22-t2v-4step.json
+++ b/tools/_comfyui/workflows/wan22-t2v-4step.json
@@ -24,7 +24,7 @@
"4": {
"class_type": "VAELoader",
"inputs": {
- "vae_name": "wan2.2_vae.safetensors"
+ "vae_name": "wan_2.1_vae.safetensors"
}
},
"5": {
diff --git a/tools/video/comfyui_video.py b/tools/video/comfyui_video.py
index e9bb0333..409264f5 100644
--- a/tools/video/comfyui_video.py
+++ b/tools/video/comfyui_video.py
@@ -57,7 +57,7 @@ _REQUIRED_MODELS_T2V = [
*_REQUIRED_MODELS_COMMON,
"wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors",
"wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors",
- "wan2.2_vae.safetensors",
+ "wan_2.1_vae.safetensors",
"wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors",
"wan2.2_t2v_lightx2v_4steps_lora_v1.1_low_noise.safetensors",
]