Thank you for using Locally AI! Here’s what’s new in this update: • Improved iPhone Duo support • Added the new LFM 3 models by Liquid AI • Enabled image understanding with Bonsai 2 models • Faster LM Link for a smoother experience with your remote models • Various bug fixes and performance improvements throughout the app

Thank you for using Locally AI! Here’s what’s new in this update: • Added the new Ternary Bonsai 2 models • Added support for M5 Neural Accelerators, with up to 4× faster prompt processing • Edit your messages and retry responses • Faster LM Link for a smoother experience with your remote models • Various bug fixes and performance improvements throughout the app

Thank you for using Locally AI! # Ready for the latest OS We’ve updated Locally for the latest OS releases, with support for the new Apple Foundation Models. Here’s what’s new in this update: • Apple Foundation Models now support vision • Added MiniCPM 5 (1B and 2B) • Added LFM 2.5 VL (3B) • Improved LM Link connection reliability and reconnection • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet the new Locally look - We’ve refreshed Locally with a new logo, a new side panel for quicker navigation on iOS, and more Liquid Glass throughout the app. Here’s what’s new in this update: • Add Gemma 4 E2B and E4B Mobile • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet the new Locally look - We’ve refreshed Locally with a new logo, a new side panel for quicker navigation on iOS, and more Liquid Glass throughout the app. Here’s what’s new in this update: • Add LFM 2.5 (2.6B) model • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet the new Locally look - We’ve refreshed Locally with a new logo, a new side panel for quicker navigation on iOS, and more Liquid Glass throughout the app. Here’s what’s new in this update: • Add Bonsai 27B on supported devices • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet the new Locally look - We’ve refreshed Locally with a new logo, a new side panel for quicker navigation on iOS, and more Liquid Glass throughout the app. Here’s what’s new in this update: • LM Link can now be used on devices not supporting MLX • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet the new Locally look - We’ve refreshed Locally with a new logo, a new side panel for quicker navigation on iOS, and more Liquid Glass throughout the app. Here’s what’s new in this update: • New side panel • More Liquid Glass • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet the new Locally look - We’ve refreshed Locally with a new logo, a new side panel for quicker navigation on iOS, and more Liquid Glass throughout the app. Here’s what’s new in this update: • New side panel • More Liquid Glass • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet LM Link - Link to LM Studio and run larger models remotely from your computer. End-to-end encrypted. Here’s what’s new in this update: • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet LM Link - Link to LM Studio and run larger models remotely from your computer. End-to-end encrypted. Here’s what’s new in this update: • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet LM Link - Link to LM Studio and run larger models remotely from your computer. End-to-end encrypted. Here’s what’s new in this update: • Add LM Link • Faster model loading • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Gemma 4* Google’s most intelligent open models to date — built for reasoning, image understanding, and multi-turn chat. * Gemma 4 is only available on supported devices with enough RAM. Here’s what’s new in this update: • Add pinned conversations • Add "New chat" button in conversations list • Improve iPad app in Windowed Mode • Fix chat scrolling and jump on new message • Improve markdown styling • Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Gemma 4* - Google’s most intelligent open models to date — built for reasoning, image understanding, and multi-turn chat. * Gemma 4 is only available on supported devices with enough RAM. Here’s what’s new in this update: - Add LFM 2.5 VL (450M) - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Gemma 4* - Google’s most intelligent open models to date — built for reasoning, image understanding, and multi-turn chat. * Gemma 4 is only available on supported devices with enough RAM. Here’s what’s new in this update: - Add the new Ternary Bonsai 8B model — a larger and smarter Bonsai model - Improve model recommendations - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Gemma 4* - Google’s most intelligent open models to date — built for reasoning, image understanding, and multi-turn chat. * Gemma 4 is only available on supported devices with enough RAM. Here’s what’s new in this update: - Performance improvements for Gemma 4 - Fix custom instructions for Gemma 4 - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Gemma 4 - Google’s most intelligent open models to date — built for reasoning, image understanding, and multi-turn chat. Here’s what’s new in this update: - Improvements to prevent crashes - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Gemma 4 - Google’s most intelligent open models to date — built for reasoning, image understanding, and multi-turn chat. Here’s what’s new in this update: - Add Gemma 4 E2B on more devices - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Gemma 4 - Google’s most intelligent open models to date — built for reasoning, image understanding, and multi-turn chat. Here’s what’s new in this update: - Add Gemma 4 E2B and E4B models on supported devices - Add LFM 2.5 350M model - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet Bonsai - Run PrismML's ultra-efficient 1-bit 8B parameters model locally — built for fast on-device performance. Here’s what’s new in this update: - Various bug fixes and performance improvements

Thank you for using Locally AI! # Rediscover Apple Foundation - Use Apple’s latest on-device model — now with better answers and improved instruction-following. Here’s what’s new in this update: - Various bug fixes and performance improvements

Thank you for using Locally AI! # Rediscover Apple Foundation - Use Apple’s latest on-device model — now with better answers and improved instruction-following. Here’s what’s new in this update: - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet The New Qwen 3.5 Models - Powerful models with advanced vision understanding and hybrid reasoning. Here’s what’s new in this update: - Performance improvements for Qwen 3.5 models - Major UI performance improvements - Various bug fixes and performance improvements

Thank you for using Locally AI! # Meet The New Qwen 3.5 Models - Powerful models with advanced vision understanding and hybrid reasoning. Here’s what’s new in this update: - Add Qwen 3.5 models - Better download progress tracking - Interface improvements - Various bug fixes and performance improvements

Thank you for using Locally AI! # More File Types Supported - You can now attach more file types, including CSV, TXT, and code files, to analyze, summarize, and more. Here’s what’s new in this update: - Add support for more file types - Add file attachments to shortcut - UX improvements - Various bug fixes and performance improvements