Wan2.2-VACE-Fun Model Setup Guide
a. Model Links and Storage Locations
Required Files:
| Name | Storage Size | Hugging Face | Model Scope | Description |
|---|---|---|---|---|
| Wan2.2-VACE-Fun-A14B | 64.0 GB | 🤗Link | 😄Link | Control weights for Wan2.2 trained using the VACE scheme (based on the base model Wan2.2-T2V-A14B), supporting various control conditions such as Canny, Depth, Pose, MLSD, trajectory control, etc. It supports video generation by specifying the subject. It supports multi-resolution (512, 768, 1024) video prediction, and is trained with 81 frames at 16 FPS. It also supports multi-language prediction. |
Storage Location:
📂 ComfyUI/
├── 📂 models/
│ └── 📂 Fun_Models/
| └── 📂 Wan2.2-VACE-Fun-A14B/
b. ComfyUI Json Workflows
1. Chunked Loading (Recommended)
Wan2.2-VACE-Fun Image to Video
Wan2.2-VACE-Fun Text to Video and Control
Wan2.2-VACE-Fun Text to Video and Control with Reference
Wan2.2-VACE-Fun Subjects to Video
Wan2.2-VACE-Fun Subjects to Video with 4 steps
2. Full Model Loading (Optional)
Wan2.2-VACE-Fun Image to Video
Wan2.2-VACE-Fun Text to Video and Control