Changelog

The Realtime Agent WebSocket now supports a text-only session: open the connection with ?mode=chat to run a conversation with no STT, TTS, or audio. The client sends user turns via input_text.send and receives the agent’s replies as transcript events. It works for any agent and is ideal for non-browser/server-side clients and automated prompt testing.

The Smallest MCP server exposes this as the new chat_with_agent tool — it holds a scripted text conversation and returns the full transcript, so an AI editor can run a build → test → refine prompt loop without placing test calls.

→ Realtime Agent reference