Notifications
Unlock the full power of local AI — zero latency, complete privacy,
and deep OS integration. Free and open source.
Direct localhost connection to Ollama, LM Studio, and local models. No network round-trips.
Chat history stays in local SQLite. API keys stored in system keychain. Nothing leaves your machine.
Global shortcut (Cmd+K), system tray, native notifications, and file drag & drop.
Native Metal / CUDA support for blazing fast local inference on Apple Silicon and NVIDIA GPUs.