OpenAI's current speech-to-speech realtime model. Served via the gateway's /v1/realtime WebSocket endpoint with text and audio input/output and function calling.
LLM Gateway routes requests to the best providers that are able to handle your prompt size and parameters.