- Smoothed chat UX by fixing a stuck “waiting” bubble case and shortening mini bubble durations.
- Increased character stability with better avatar loading, animation syncing, and collider handling.
- Improved library reliability and visuals, including safer defaults and better performance on older AMD systems.
- Updated Local LLM framework to allow for the use of model models like Qwen 3
- Strengthened Local LLM usability with more reliable model setup/download behavior and graceful fallback when needed.
- Applied general UI polish across settings and core scenes for a cleaner, more consistent experience.
Update 1.2.4 03-04
Source: Steam Community (opens in a new tab)