ComfyUI-WanVideoWrapper  by kijai

ComfyUI nodes for advanced video generation

Created 8 months ago
5,127 stars

Top 9.7% on SourcePulse

GitHubView on GitHub
Project Summary

This repository provides ComfyUI custom nodes for integrating WanVideo and related video generation models, targeting users who want to experiment with cutting-edge video AI capabilities within the ComfyUI ecosystem. It offers a faster development path for new models and features not yet supported natively, enabling rapid testing and exploration.

How It Works

The wrapper acts as a bridge, allowing ComfyUI to utilize various video generation models, including WanVideo, SkyReels, ReCamMaster, and others. It simplifies the integration of these models by providing dedicated nodes, abstracting away complex loading and configuration steps. This approach facilitates quicker iteration on new model implementations and experimental features compared to contributing directly to the ComfyUI core.

Quick Start & Requirements

  • Install by cloning the repository into your ComfyUI custom_nodes folder.
  • Install dependencies: pip install -r requirements.txt or use the provided portable install script.
  • Models need to be placed in specific ComfyUI subdirectories (e.g., models/diffusion_models).
  • Supports GGUF models and native ComfyUI text encoders/clip vision loaders.
  • Links: Hugging Face Models

Highlighted Details

  • Supports a wide array of video models including SkyReels, WanVideoFun, ReCamMaster, VACE, Phantom, ATI, Uni3C, MiniMaxRemover, MAGREF, FantasyTalking, MultiTalk, and EchoShot.
  • Offers fp8 scaled models for potentially improved performance and reduced VRAM usage.
  • Includes patcher nodes for integrating experimental models into native ComfyUI workflows.
  • Demonstrates context window capabilities with large frame counts and VRAM efficiency.

Maintenance & Community

This project is described as a personal sandbox and is perpetually in a "work in progress" state, indicating a focus on rapid development and experimentation rather than long-term stable maintenance. There are no explicit mentions of community channels or a roadmap.

Licensing & Compatibility

The repository's licensing is not explicitly stated in the README. Users should verify compatibility for commercial use or integration with closed-source projects.

Limitations & Caveats

The project is explicitly labeled as "WORK IN PROGRESS (perpetually)" and prone to issues due to the developer's self-described lack of coding experience and the complexity of ComfyUI. It is not intended to compete with or replace native ComfyUI workflows but rather to serve as an experimental platform.

Health Check
Last Commit

9 hours ago

Responsiveness

Inactive

Pull Requests (30d)
19
Issues (30d)
222
Star History
472 stars in the last 30 days

Explore Similar Projects

Starred by Jiaming Song Jiaming Song(Chief Scientist at Luma AI) and Yineng Zhang Yineng Zhang(Inference Lead at SGLang; Research Scientist at Together AI).

LightX2V by ModelTC

6.3%
741
Video generation inference framework for efficient synthesis
Created 7 months ago
Updated 9 hours ago
Starred by Yineng Zhang Yineng Zhang(Inference Lead at SGLang; Research Scientist at Together AI), Yaowei Zheng Yaowei Zheng(Author of LLaMA-Factory), and
1 more.

FastVideo by hao-ai-lab

1.2%
3k
Framework for accelerated video generation
Created 1 year ago
Updated 2 days ago
Feedback? Help us improve.