Multimodal AI

ComfyUI

Comfy-Org/ComfyUI

A node-based interface and execution engine for building image and other generative AI workflows.

★ 136.7KStars
⑂ 16.3KForks
4300Open issues
PythonLanguage
GPL-3.0License
Q84Editorial score

Overview

ComfyUI represents model pipelines as connected nodes, making generation workflows inspectable, reusable, and extensible. The project provides a visual interface, backend execution, and an API for integrating workflows into other applications.

Key features

  • Visual node graph for model pipelines
  • Reusable and shareable workflows
  • Extensible custom nodes
  • Backend and API access

Requirements, installation and quick start

Follow the official installation guide for your operating system and GPU setup. Requirements vary by model and acceleration backend.

Usage

Load or create a workflow graph, select compatible checkpoints and inputs, then queue execution in the interface or submit it through the API.

Model compatibility and use cases

Supports multiple image-generation model families through built-in and community nodes; compatibility depends on node versions and model files.

License and risk notes

The repository declares GPL-3.0. Review the license obligations before redistributing modified versions or bundling it in a product.

Release and maintenance

See the official repository for current releases and changes.

UniRL

Tencent-Hunyuan/UniRL

★ 854Python

diffusers

huggingface/diffusers

★ 34.5KPython

GLM-V

zai-org/GLM-V

★ 2.4KPython