A few bug fixes.

- Add Qwen3-VL and Ministral 3 support - Bug fixes

- Adds Shortcuts - Made some UI enhancements

- Added integration with Palshub.ai - Performance enhancements and numerous other improvements under the hood

- Added LFM2-VL models support

- Import/Export Pals: Easily back up or share your pals! - GPT-OSS Support: Enabled! May your device have enough RAM... and courage. - Report Content: Now you can report inappropriate content. - Bug Fixes.

- Added support for LFM2 models

- Added thinking control

- Added SmolLM3 support

- Added support for ERNIE models - Added a copy button for markdown code blocks - Fixed a bug affecting mostly Gemma models

- Added Gemma 3n E4B support - Added code copy feature - General enhancements and improvements

- Introducing Lookie Pal: your new vision-enabled companion! - Performance improvements and bug fixes.

- General bug fixes and performance improvements.

- Revamped Thinking Bubble: Enjoy a sleek new design for the thinking bubble. - Prompt Configuration Update: You can now control whether reasoning tokens are included in prompts via a new parameter. - Chat Session Import/Export: Easily back up or share your conversations with the new export and import functionality.

- Added support for Qwen3 - Improved Japanese translations

- Hugging Face Token Authentication: Secure access to gated models using your personal Hugging Face token. - Localization Support: Now available in Japanese and Chinese, in addition to English — enjoy the app in your preferred language!

- Completion settings now can be set either for a chat session or as preset - Added App Info screen with user feedback option

- Now available on iPad! - Added support for EXAONE-Deep with reason rendering

Adding support for Gemma 3.

- Introducing Pals! Create roleplay or assistant pals to enhance your chat experience. - Revamped Chat UI with improved design, updated model settings, and a clearer chat header displaying the active model and session title. - Session Management Upgrades – Easily rename, delete, or duplicate sessions. - Enhanced Chat Menu featuring chat generation settings, model selection, and session controls. - Screen Awake Mode – Keeps the screen active during inference and deactivates when idle for better usability.

Added: DeepSeek R1 model support Improved: A few UI elements. Squashed: Minor bugs, major vibes.

- New Benchmarking Page: Test and compare performance metrics with ease using our newly added benchmarking feature. - Sync with the latest llama.cpp. - Advanced Settings: Fine-tune your experience with exposed settings for Flash attn, batch, ubatch, and cache_type_kv, now available in both the UI and benchmarking results. - Bug Fixes: We've squashed a few minor bugs. - UI Enhancements: Enjoy an improved interface with subtle yet effective tweaks.

• Enhanced chat features: Edit messages and retry AI generations • Fresh new UI theme for a better user experience • Added background download option

Hugging Face Public Hub Integration: Effortlessly search for GGUFs on Hugging Face directly within the app. Bookmark your favorites or download directly from the app.

Added SmolLM to the default list of models.