comfyui-ollama

Why we included this project

ComfyUI users who want language model output inside their visual workflows can install these custom nodes and get a direct bridge to a running Ollama server, no need to leave the editor or switch tools. The nodes are written in Python and JavaScript, handling connection setup, chat history, and API options, so the only real requirement is an Ollama instance reachable from the machine running ComfyUI. The chat node keeps multi-turn context natively and lets you chain models, which means one model's reply can feed straight into the next step of a generation pipeline. For artists automating prompt generation or batching operations, that integration is the main appeal; the LLM plumbing is handled for you and the whole workflow stays inside ComfyUI.