v4.6.1
📦 localaiView on GitHub →
✨ 2 features🐛 6 fixes
Summary
This patch release focuses on stability and resource management, introducing a new API endpoint for model capabilities and fixing several issues related to authentication, UI presentation, and VRAM utilization for context and embeddings.
✨ New Features
- Added GET /v1/models/capabilities endpoint to the API.
- Implemented native Prometheus metrics for agent chat runs.
🐛 Bug Fixes
- Fixed OIDC/OAuth user-info failures by logging the real cause.
- Centered the home empty-state wizard in the UI.
- Prevented persistence of request-scoped reasoning_effort as an operator disable.
- Scoped generated-content and upload directories to the current user during startup.
- Capped auto-derived context to fit available VRAM.
- Capped single-pass embedding batch size to fit VRAM in llama-cpp operations.