This is my version of OpenAI's Realtime Console. I've added a few features to make it more useful for my needs and work with speaches, such as:
- MCP support
- Keyboard shortcuts
- Conversation persistence
- WebRTC data channel message fragmentation
- More comprehensive session configuration
- Few other things I don't remember at the top of my head...
I've vibecoded most of this code, so it may not be the most efficient or elegant solution. Any contributions or suggestions for improvement are welcome!