Grok Voice Think Fast 2.0 now available on AI Gateway

Vercel added xAI's Grok Voice Think Fast 2.0 to AI Gateway as a speech-to-speech model: it accepts audio and returns spoken replies. The release notice says the model reasons while speaking and can begin tool calls earlier in a response.

For application developers, the integration uses the AI SDK's realtime interface. Its example creates short-lived access credentials on the server so the long-lived API key is not delivered to the client. The announcement describes a route to building responsive voice agents, with the responsiveness improvements presented as vendor claims rather than independently measured latency results.

“The model reasons in parallel with speech, so it can think through a query while talking without adding latency.”

“It has also been trained to use fewer reasoning tokens than before, so tool calls fire sooner, often before the end of the agent's first sentence.”

“Transcription holds up in real-world conditions, including background noise and telephony compression.”