The model used is gpt-4o-realtime-preview-2024-12-17. Is there anything else that needs to be configured?