Skip to main content

Ollama

Ollama runs LLMs locally on your machine. Free, private, no API keys.

Setup​

# macOS
brew install ollama
ollama serve

# Pull models
ollama pull gemma3:4b        # fast, 4GB
ollama pull llama3.2         # strong reasoning
ollama pull qwen2.5-coder    # optimized for code

Config​

{
  "activeProvider": "ollama",
  "providers": {
    "ollama": {
      "apiKey": "ollama",
      "baseUrl": "http://127.0.0.1:11434/v1",
      "model": "gemma3:4b"
    }
  }
}

Auto-detection​

If ollama is not in your config, ak-coder tries http://127.0.0.1:11434/api/tags at startup and picks the first available model from a preference list: gemma4:31b-cloud, gemma4:12b-mlx, gemma4:latest.