Ollama

Use this when Provider is set to ollama (local/self-hosted inference).

Configure in app

1. Open Settings > AI Model.

2. Set Provider to ollama.

3. Set:

- ollamaBaseUrl (default: http://localhost:11434)

- ollamaNumCtx (optional context window override)

4. Choose model names that are installed in your Ollama runtime.

Setup and model install

  • Install and run Ollama:

- Ollama download and docs

  • Pull models locally (example):

- ollama pull llama3.1

Notes

  • No cloud API key is required for local Ollama.
  • Ensure the app can reach the host/port in ollamaBaseUrl.