Connect Ollama to Cursor

Last verified: 2026-08-24 · Example model: llama3.1

1. Install Ollama and pull a model

Follow an install guide for your OS, then pull a model:

• Ollama on Windows
• Ollama on macOS
• Ollama on Linux

ollama pull llama3.1

Pick another model from our Ollama hub or a per-model run guide.

2. Confirm Ollama is running

1. Run ollama list to confirm your model is downloaded.
2. Ollama serves an OpenAI-compatible API at:

http://localhost:11434/v1

3. Connect Cursor to localhost

1. Open Cursor Settings (Ctrl+Shift+J / Cmd+Shift+J).
2. Go to Models.
3. Enable Override OpenAI Base URL and paste the Ollama URL above.
4. If Cursor asks for an API key, use any placeholder such as ollama.
5. Add a model with the exact Ollama tag:

llama3.1

6. Select the model in chat and send a test prompt.

Troubleshooting

• Connection refused — start Ollama and run ollama serve if needed.
• Model not found — the name must match ollama list exactly.
• Slow or out of memory — try a smaller model or check VRAM on the model card.

← Local guides · Ollama hub