Loading...
- v7.0iOS
- Dreamshaper XL Lightning (Low memory model) - Support for Bonsai 8B & Gemma 4 - Chat enhancements - Bug fixes
- v6.7iOS
- v6.6iOS
- Dreamshaper XL crash fix - Bug fixes
- v6.5iOS
- iCloud sync fix - Bug fixes
- v6.4iOS
- Dreamshaper XL crash fix
- v6.3iOS
- Crash and bug fixes
- v6.2iOS
- Added Qwen3.5 GGUF models - Added support for Qwen3.5 GGUF models - Almost doubled MLX inference - Bug fixes
- v6.1iOS
- Added support for Hebrew - Audiobook localization - Audiobook improvements - Added support for Hebrew - 4x Upscale improvements - Thinking model enhancements - Added support for RRDB, HAT, Official HAT, DAT, or ATD upscale model architecture - Critical bug fixes and enhancements
- v6.0iOS
- Audiobook converter: you can now convert epub to an audiobook using beautiful KokoroTTS voices and Hifigan model. (English only) - Updated 4x upscaler - Added Support for Qwen3.5 vision models - Bug fixes and improvements
- v5.0iOS
- Enhanced web search - On-device Qwen3TTS (beta) - Bug fixes and improvements
- v4.0iOS
- 27-Language Localization - Lekh AI now speaks your language. The entire app - every menu, setting, button, and message - is localized into 28 languages. Set your App Language for the UI and a separate Response Language for AI replies. - Improved Model Catalog - Updated model library with new downloads and better management. Browse, download, and configure 44+ AI models directly from the app. - Bug Fixes & Performance - General stability improvements, updated dependencies, and UI refinements across the app.
- v3.0iOS
- Updated schedule reminders/prompt runs - 4x Upscaling - Chat translation (setting framework for entire app transalation) - Bug fixes
- v2.6iOS
- In-app purchase fix. We apologize for any disruptions to new users. - Handsfree TTS mode: Enable the mode and say "Hey Lekh" - Chat branching - Remote models - Webpage analysis - Web search using duckduckgo - AI translation - Flash cards - Hifigan TTS - Stability fixes and more
- v2.5iOS
- Lekh AI is now free to download - New users get a 3-day free trial. - Early supporters retain full lifetime access – thank you for believing in on-device AI - Added support for External Servers. You can connect to servers like LM Studio and have seamless conversations with premium models. - Bug fixes and performance fixes
- v2.0iOS
- Added support for llama GGUF models - Better RAG memory - TTS optimizations - Removed consistent uncensored models warning - Bug fixes and performance fixes
- v1.3iOS
- Qwen3-TTS and voice cloning (for Advanced users) - Added back link to onboarding - Local server now persists on app closes - More descriptive models - Bug fixes - Perfomance improvements
- v1.2iOS
- Custom personas - AI Image editing with SD 1.5 - Black and white image colorizer - 3D Splat generator and viewer - Photo to 3D - Object remover - View and fine tune RAG memories - HTML artifacts, Code blocks, and chat view enhancements - Better RAG integration - Background model downloading - Chat summarization - Ask about your book progress in chat - Fine tune memories - More Siri shortcuts - Performance and memory management improvements - Several bug fixes - Significantly decreased app size by offloading FastVLM and making it optional
- v1.1iOS
- Bug fixes - Performance improvements - Library beta addition - Image generation improvements - RAG improvements - Floating menu