2.2.2
2026-01-17
-Add support for Granite 3.1 3B A800M Instruct
-Reworked summaries to support rolling sumamries improving output while using small models
-Added group/user space to projects list and details screen
-Added support for mermaid diagram rendering should a mermaid block appear in README.md
-Fixed an issue causing the keyboard to cover the text field when typing messages to one of the chatbots
-Fixed an issue causing LLM settings to show when "No AI" is selected
-Fixed an issue causing non-ascii characters to show onscreen in LLM output
-
2.2.1
2026-01-08
Added support for on-device LLM inference
Added on-device support for Qwen 2.5 7B, Qwen 3 8B, Phi 4 Mini 4B, Gemma 3 4B, and TinyLlama 1.1B models
Provided the option to configure a routing model (e.g. FunctionGemma) for dynamic tool selection
Tweaked the settings pages for better usability
Added a toggle to switch Lab Assistant on and off
Added an option to totally disable AI functionality"
1.1
2025-08-07
Updated token validation logic to allow for formats outside of GitLab CE
Fixed a bug causing the breadcrumb to overflow off of the screen while viewing file contents when the branch dropdown was too long
Improved settings saving logic for Ollama settings