jieke360 / ComfyUI-Stable-Video-Diffusion

ComfyUI nodes for Stable Video Diffusion

ComfyUI Stable Video Diffusion

Easily use Stable Video Diffusion inside ComfyUI!

• Installation
• Node types
• Example workflows
  • Image to video
  • Image to video generation (high FPS w/ frame interpolation)

Need help? Join our Discord!

1. Installation

cd ComfyUI/custom_nodes/
git clone https://github.com/thecooltechguy/ComfyUI-Stable-Video-Diffusion
cd ComfyUI-Stable-Video-Diffusion/
pip install -r requirements.txt

2. Download models into ComfyUI/models/svd/

• svd.safetensors - Download
• svd_image_decoder.safetensors - Download
• svd_xt.safetensors - Download
• svd_xt_image_decoder.safetensors - Download

Node types

• SVDModelLoader
  • Loads the Stable Video Diffusion model
• SVDSampler
  • Runs the sampling process for an input image, using the model, and outputs a latent
• SVDDecoder
  • Decodes the sampled latent into a series of image frames
• SVDSimpleImg2Vid
  • Combines the above 3 nodes above into a single node

Example workflows

Image to video

https://comfyworkflows.com/workflows/5a4cd9fd-9685-4985-adb8-7be84e8636ad

Image to video generation (high FPS w/ frame interpolation)

https://comfyworkflows.com/workflows/bf3b455d-ba13-4063-9ab7-ff1de0c9fa75