New
v1.0.1 — First public release
- Spins up a local server that speaks the OpenAI
/v1/chat/completionsAPI, backed by the new macOS 26 Apple Intelligence models. - Everything runs on-device (no network, no keys).
- Streaming responses (
"stream": true) already supported. - Basic health endpoints:
/health,/v1/models.
Point any OpenAI-compatible client at http://127.0.0.1:11535 and chat away!