ComfyUI (Core Repository)
The core open-source ComfyUI engine — a modular node-based interface for generating images, video, 3D, and audio using diffusion models.
ComfyUI is the core open-source node-based engine for generative AI, developed by Comfy Org. It provides a modular graph interface where users can build custom pipelines by connecting nodes for model loading, sampling, image processing, video generation, 3D creation, and audio production.
The repository is the primary development hub for the ComfyUI ecosystem, with contributions from the community and active daily development by the core team. It supports a vast library of custom nodes via the ComfyUI-Manager and the official Registry.
Related Sites
Comfy CloudRun ComfyUI online with zero setup — pre-loaded models, pre-installed custom nodes, and concurrent jobs powered by NVIDIA RTX 6000 Pro GPUs. No local hardware required.
Comfy Cloud MCPThe official Comfy Cloud MCP server repository — connect AI agents to Comfy Cloud for image generation, model search, and workflow management via the Model Context Protocol.
Comfy MCPDrive ComfyUI from any AI agent via the Model Context Protocol. Generate images, video, audio, and 3D from Claude Code, Claude Desktop, Cursor, and other MCP-compatible clients.
Comfy OrgThe official brand site for Comfy — the open-source node-based engine for generative AI. Download Desktop, browse workflows, and access the ecosystem.