v1.45 - Update LlamaCPP framework - Restructure the framework for future updates - Update LLM model to Qwen3.5-4B - Minor UI update to the paste button in MacOS

v1.44 - Update LlamaCPP framework - Resolve bug when loading previous session

v1.43 - Update LlamaCPP framework - Update LLM model to Qwen3-VL-4B-Instruct-Q6

v1.42 - Update LlamaCPP framework - Update LLM model to Qwen3-4B-Instruct-2507 - Remove the "think toogle"

v1.41 - NEW feature: connect to MCP HTTP services - Update LLM model to Qwen3-8B-IQ3_XS - Update LlamaCPP framework - Optionally reload from "Settings" with updates - NOTE: URL requests are part of "Enable External Tools" - UI will be in dark mode despite system configuration

v1.40 - NEW feature: AltBot can make search queries NOTE: it requires a third party SearXGN service - URL requests are listed as a tool - Enable/disable model thinking - Localization for Spanish and French for these new features - Refactor code to easily add more characteristics in the future - Update model to Qwen3-4B-Q6_K

v1.36 - Update LLM model to DeepSeek-R1-Distill-Qwen-7B-IQ3_XS - Show some markdown elements as created from the model - Add a distinctive message style for a the "Thinking" phase - In mac, use "cmd + shift + c" to clear the current context - Change settings to always work correctly in iPhone - Add the "Temperature" to settings.

v1.35 - Update LLM model on Mac: granite-3.0-8b-instruct-IQ3_XS - Update LLM model on iPhone/iPad: granite-3.0-3b-a800m-instruct-IQ4_XS - AltBot could run on iPad Pro 2nd Gen, but performance is really slow - Resolve issue on Mac for the chat when tokens were not followed - If user moves the UI in iPhone, it stops following until the next message NOTE 1: AltBot ran on iPad Pro (2ndGen) with 4GB with this update. But performance was substandard. NOTE 2: This is the last update to support older versions to MacOS 15 Sequoia. Next year this will be a requirement for newer versions On iPhone/iPad, it will require iOS 18 for newer versions.

v1.34 - Update LLM model: Llama-3.1-Storm - Add full screen hotkey: CMD + CTRL + F - Settings is within the chat window (mac) - Enable Chat export

v1.33 - Add French and Spanish translations - Allow users to change the maximum input tokens in Settings - Your session files are cyphered - Experimental: Allow AltBot to make outgoing HTTPS requests to read webpages NOTE: Experimental features could change between versions NOTE #2: Previous session data will be ignored

v1.32 - Make sure sessions work as expected - Use context shift to maintain consistency with the conversation - Update base model to Llama3.1

v1.31 - Settings view - Credits - Bump up to macOS 14 as minimum requirement - Bug fix: follow tokens loading with scroll view

Changes: - Loading Animation - Update LlamaCPP - Update model to Llama-3-Smaug - Add "Save/Clear" buttons to store/remove session data