xAI's Grok Voice 3.0 targets phone support at $0.08 a minute

xAI released Grok Voice 3.0 on October 14. It is a speech-to-speech model, meaning it listens and talks back directly instead of passing audio through separate transcription and text-to-speech steps. Developers reach it through xAI’s Voice Agent API.

The price stays at $0.08 per minute, about $4.80 an hour, the same as Grok Voice 2.0. xAI says the first audio comes back in under a second. Agents already set to grok-voice-latest switched over on their own.

xAI says the model powers voice in Tesla vehicles, the Grok Bot’s voice mode and Starlink’s customer call centers. One report says about 70% of Starlink calls end without a human. The company also claims the top score on Artificial Analysis’s Speech-to-Speech Index, 86.9 against 81.3 for the previous version.

The catch: The benchmark numbers and the call-resolution figure come from xAI, and the coverage we found has no independent test.

Should you try it?

• Maybe yes: you take lots of routine calls (hours, bookings, order status) and can afford a developer or a no-code tool that connects to the API.
• Probably not yet: you want a ready-made answering service, or your calls are sensitive and need a human early.

Prompts to test it

You are the phone assistant for [business type, e.g. a dental clinic]. Answer questions about hours, location and booking. If a caller sounds upset or asks about [sensitive topic], offer to pass them to a person. Keep every answer under two sentences.