🏄‍♂️ refactor: Optimize Reasoning UI & Token Streaming

[danny-avila]

Summary

• Fixed a rendering issue with long generations (10k+ chars), typical of deepseek-reasoner that would slow down the entire app. after a certain length.
• Optimized tokens in transit, introducing data improvements similar to Agents.
  • Besides Agents, only applies to azure, openai, and custom endpoints
• Removed message caching unless requested by Audio Request
• Updated label text to show “Thinking...” during message submission
• Provided a “Show Thinking” option in settings for controlling the default state
• Added a toggle to display or hide reasoning content by default

Other Changes

• Fixed latest message auto-playback not playing on chrome
  • https://stackoverflow.com/questions/50490304/how-to-make-audio-autoplay-on-chrome-
• Refactored the TTS code so audio requests happen after message streaming completes
• Fixed edge cases when editing artifacts

Checklist

• My code adheres to this project's style guidelines
• I have performed a self-review of my own code
• I have commented in any complex areas of my code
• My changes do not introduce new warnings
• Local unit tests pass with my changes