When enabled, requests are sent to reasoning model with reasoning_effort=high. No temperature param sent.
Auto-scroll on new messages
Scroll to bottom after AI responds
Show message timestamps
Display time of each message
Markdown rendering
Render markdown in AI responses