Fast Models: Faster Responses, Less Waiting

Features

• Fast models: Get faster responses with less waiting for a smoother conversation experience.

Improvements

• Added a notice that the Quest window will be retired soon.
• The Q&A panel in the conversation flow can now be expanded or collapsed as needed.
• Improved performance in long conversations and large edits, with fewer freezes and crashes.
• Improved the SSH remote experience with more reliable connection and setup.
• Improved how model requests are sent, so large requests go out faster and responses stay more stable on weak networks.
• Adjusted how often automatic incremental indexing runs, so frequent indexing uses fewer resources.