Overview

ComfyUI (Comfy) is a professional visual AI creation engine that gives you granular control over every model, parameter, and step in your generation pipeline. Instead of a single prompt box, you build workflows on an infinite node canvas—chaining checkpoints, samplers, LoRAs, ControlNets, upscalers, and video models into reusable graphs. The free desktop app bundles a stable ComfyUI build with automatic Python dependency management, the built-in ComfyUI-Manager for installing custom nodes, and auto-updates, so you can run state-of-the-art open models locally on Apple Silicon with Metal acceleration. App Mode hides the graph behind a simple form when you just want to run a workflow, and thousands of community templates give you a starting point for image generation, editing, and video.

Pricing: Free & open source (optional paid Comfy Cloud)

Minimum macOS: 12.3 (Monterey)

Architecture: Apple Silicon

Key Features

Node-based workflow builder on an infinite canvas
Every model, parameter, and step is visible and adjustable
App Mode for a simplified interface, with one toggle back to the full node graph
One-click desktop installer with automatic updates
Automatic Python environment and dependency management
Built-in ComfyUI-Manager for installing and updating custom nodes
60,000+ community nodes and extensions
Thousands of shareable community workflow templates
Supports text-to-image, image-to-image, inpainting, and upscaling
Video generation with open video models
ControlNet, LoRA, and custom checkpoint support
Runs locally on Apple Silicon with Metal (MPS) acceleration
Portable and CLI installation options alongside the desktop app
Optional Comfy Cloud and API for running workflows remotely
Fully open source with no feature gates or pro tier