Release Notes v0.11.3
Custom Providers
Add your own OpenAI-compatible providers (LM Studio, vLLM, Ollama OpenAI endpoint, etc.) via Settings → Provider. Custom providers appear with ★ prefix and support optional API keys for local servers.
Qwen Regional Endpoints
Choose between International (Singapore), China (Beijing), or US (Virginia) endpoints in Settings → Advanced → Provider Settings. API keys are region-specific.
Model Updates
- Comprehensive audit of all 16 providers with latest models
- New tier system (reasoning/flagship/standard/fast/ultrafast) for model organization (for later use in Actions)
- Fixed Kimi model IDs for native Moonshot API (K2 series)
Bug Fixes
- Fix format string crashes with some locales (Thanks to user CLEMENTINE2021) issue #27
- Fix translation file syntax errors
- Providers and models now sorted alphabetically in all menus
Full Changelog: v0.11.2...v0.11.3