Installing Ollama

Ollama is a separate program from ComfyUI. It must be installed and actively running in the background, or every ComfyUI generation fails the moment it reaches the prompt-building nodes. This is the single most common reason a fresh setup does not run.

Steps

1. Install Ollama for the target operating system from ollama.com.

2. Pull the model the workflow expects:

   ollama pull llama3.1:8b

1. Leave Ollama running. The ComfyUI workflow communicates with it at http://127.0.0.1:11434. That address is baked into the workflow, so the service has to be live at that port whenever you generate.

Confirm it is running

Open http://127.0.0.1:11434 in a browser. A short “Ollama is running” response means the service is live and ComfyUI can reach it.

If generation in ComfyUI fails at the prompt nodes, check this first. A stopped Ollama service is the most likely cause before you start debugging the workflow itself.