Overview

ComfyUI is a modular GenAI inference engine that can be run as a server, accessed via API, extended with custom nodes, and managed from the command line. Choose your path below.

​

Build on ComfyUI

Run ComfyUI workflows from your own application and get the results back. The official SDKs sit on a versioned HTTP API, and the same code works against Comfy Cloud or a ComfyUI instance you host yourself. See also: Design Notes · Comfy API v2 Reference · APIs Overview

​

Deploy ComfyUI as a Server

Run ComfyUI in your own environment and expose it as an API endpoint. Learn about the WebSocket message protocol, available routes, and execution modes. See also: API Examples · Partner Node API Integration

​

Comfy API

The public API for running ComfyUI headlessly has evolved over time. For new integrations, use Comfy API v2. It is supported on open-source ComfyUI, Comfy Cloud, and the new Serverless API. Older versions include the v1 Cloud API, which is deprecated. See also: Serverless API · v1 Cloud API Overview · Getting an API Key · APIs Overview

​

Agent Tools / MCP

Connect AI agents to ComfyUI via the Model Context Protocol (MCP). Start with the hosted Cloud MCP, or use Local MCP and Comfy CLI for other setups. See also: Comfy MCP — cloud and local connections

​

Comfy CLI

Manage your ComfyUI installation, custom nodes, and workflows from the command line. Install, update, configure, and automate with ease. See also: CLI Reference · Troubleshooting

​

Custom Nodes

Extend ComfyUI with Python backends, JavaScript UI extensions, and integrations. Start with the Custom Nodes overview.

​

Registry

Package and publish custom nodes through the Comfy Registry. Start with the Registry overview or the publishing tutorial.