ComfyUI

ComfyUI

Alternative to Automatic1111 WebUI

Node-based visual AI engine and application

133k 16k
GPL-3.0
last commit 2026-03-06
Website Source
Share:

About

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.

ComfyUI website preview

Languages

Contributors30

  • Node-based workflow design
    Core

    Build complex AI generation workflows visually by connecting functional nodes on a canvas.

  • Generative AI engine
    Core

    Execute diffusion model workflows for image, video, and other media generation with fine-grained control.

  • Advanced control & customization
    Core

    Access and adjust low-level parameters of models and samplers to precisely tailor outputs.

  • Reusable & shareable workflows

    Save, load, and share entire workflow configurations for reuse across projects.

  • Multi-platform support

    Run on Windows, Linux, and macOS with support for various GPU types and CPU fallback. :contentReference[oaicite:1]{index=1}

  • Extensions & custom nodes

    Support for community-created custom nodes and extensions to expand functionality.

  • Model flexibility

    Integrate a wide range of diffusion and auxiliary models (e.g., ControlNet, LoRA).

  • Smart execution optimizations

    Efficient execution that only re-runs parts of workflow graphs that change between runs. :contentReference[oaicite:2]{index=2}

Comments Theme
Hosting
self_hosted
managed_cloud
Interface
gui
api
Install
docker_compose
slug: comfyui