xAI has released Grok Voice 3.0, its new speech-to-speech model, and it didn't raise the price. The Voice Agent API still costs $0.08 a minute, or $4.80 an hour, the same as Grok Voice 2.0. If your agent points at grok-voice-latest, it is already running the new model and you don't need to change anything.
The headline number comes from Artificial Analysis, which xAI quotes: Grok Voice 3.0 scores 86.9 on the Speech-to-Speech Index, ahead of GPT-Live-1 at 83.2, Gemini 3.8 Live at 82.6 and its own predecessor at 81.3. That index blends speech reasoning, the τ-Voice support-call test, and blind head-to-head preference and task success. xAI also cites 97% on Full Duplex Bench, which measures how naturally a model handles pauses, turn-taking and interruptions, and 71.6% on τ-Voice, which it says is comparable to GPT-Live-1 at about 40% lower latency. On ServiceNow's EVA-Bench for enterprise voice agents, the model posts the top accuracy score at 74%.
xAI says the model already handles millions of conversations a day in Tesla cars. Starlink's call centers also run on Grok Voice, taking about 20,000 calls a day, and roughly 70% of them are resolved without a human.