Highlights

• Privacy settings & analytics controls

Enhancements

• Token/speed is persisted per message
• Added chat width settings
• Improved default settings of right sidebar & input box for new users
• Improved the UX when models are deleted
• Better error handling for some cases

Bug Fixes

• Enabled OpenAI GPT-4o model’s vision capability
• Fixed auto-scrolling and performance issues in chat
• Fixed markdown component crashes and rendering issues
• Resolved max_tokens configuration preventing remote model usage
• Fixed other various UI and threading issues