• Image support for GPT 4o and Claude
• PDF support (with caching) for Claude
• Full width chat on wide screens
• Option to set LM Studio local network URL
• Faster rendering (React fixes)

Fix bugs

• Fix React mistake that was causing slowness on typing
• Optional sign-in

Fixed some weirdness with markdown rendering

Features

• Added Perplexity and Gemini Flash Thinking
• Added link previews for Wikipedia links
• Improved title generation

Bugs

• Fixed O1 not rendering bug
• Added note about enabling CORS for LM Studio models
• Fixed code block rendering
• Allow for selection of parts of an answer

You can now run local models! Download Ollama or LM Studio (or both) and go to the /models page (cmd+shift+m) to try it out.

Add bring your own API key option

Fixed O1

Tweaks

Updated styling, proxy

Home page tweaks

Settings page styling updates

Redirect to / on chat deletion

• Graph view
• Add a new model after generations
• CMD+L to focus on textarea
• Empty states

Updater fixed

CMD+index shortcut to select

Tweaks

New data model, UI improvements

Chorus is updated! Check the changelog for details.

Renamed to Chorus, design updates.

Fix new model bug.

Multichat experiment

Fix gemini

Multi chats!

Melty Makeover!

• Fix bugs

• Fix cmd key override

• Set @tool to default enabled

• Melty now suggests which agent to call
• New @tool program
• Fixed broken edit button