Introducing the Realtime API

https://openai.com/index/introducing-the-realtime-api/

Speech-to-speech over a persistent WebSocket with GPT-4o (gpt-4o-realtime-preview), automatic interruption handling, function calling. Audio $100 and $200 per 1M input and output tokens, about $0.06 per minute in and $0.24 per minute out; text $5 and $20. Later updates through general availability in August 2025.

Realtime speech-to-speech and expressive TTS for characters