Comfy-Org/Comfy-Desktop

The desktop app for ComfyUI

What it solves

Comfy Desktop removes the technical friction of installing and managing ComfyUI. It eliminates "dependency hell" by providing a self-contained environment, removing the need for users to manually manage Python, CUDA, and pip installations or deal with terminal commands.

How it works

The application acts as a manager for ComfyUI instances. It provisions isolated, GPU-ready environments that include a relocatable Python with prebuilt PyTorch and GPU wheels. It also bundles a minimal Python bootstrap with pygit2 to handle Git operations (like cloning) on machines that do not have Git installed.

Who it’s for

Users of ComfyUI who want a streamlined, one-click installation process and the ability to manage multiple independent setups without conflicts.

Highlights

  • Side-by-side installs: Run multiple independent ComfyUI setups, each with others' versions, models, and custom nodes.
  • Isolated environments: GPU-ready environments are prebuilt to avoid installation failures.
  • One-click updates: Update the core engine and custom nodes without using a terminal.
  • Snapshots and rollback: Create backups of an installation to restore it if updates or custom nodes cause breakage.
  • Existing setup migration: Import existing portable or Git-based ComfyUI installations.
  • Cross-platform: Supports Windows, macOS (Apple Silicon), and Linux.

Related

  • Project
    Tavris1/ComfyUI-Easy-Install

    A portable, one-click installer for ComfyUI on Windows that includes the EZi Desktop management tool to simplify the installation of complex AI packages and environment configuration.

  • Project
    mmartial/ComfyUI-Nvidia-Docker

    A Docker-based deployment for ComfyUI optimized for NVIDIA GPUs, providing isolated environments with support for a wide range of CUDA versions and hardware.

  • Project
    YanWenKun/ComfyUI-Windows-Portable

    A ready‑to‑run Windows package that bundles ComfyUI (a node‑based UI for Stable Diffusion) with 40+ custom nodes and 300+ pre‑compiled Python packages, letting users launch AI image/video generation instantly on an NVIDIA GPU.

  • Project
    Comfy-Org/comfy-cli

    comfy‑cli is a Python CLI that automates installing, launching, and extending ComfyUI – an open‑source generative‑media engine. It lets you install custom node packs, download models, run local or cloud workflows, and call partner image/video models directly from the terminal, with optional JSON output for AI agents.

  • Project
    Comfy-Org/ComfyUI-Manager

    ComfyUI‑Manager is a feature‑rich extension for the visual AI workflow tool ComfyUI. It adds a UI button that lets users browse, install, update, enable/disable, and remove custom nodes and model files from the official registry or community channels. The manager also provides snapshot/restore of installation states, one‑click sharing to workflow‑hosting sites, a CLI (`cm‑cli`) for automation, and security‑focused storage of its data. Configuration is handled via a simple `config.ini` and optional pip‑override files. Installation works via manual git clone, portable‑Windows script, `comfy-cli`, or a Linux venv script.