Hi everyone. I’ve been recently using the realtime API to communicate with user-prompted assistants. This communication has two separate implementations. One implementation creates a WebRTC session for client-side (browser) voice interaction and a websocket…