Back to stories
Jun 3, 2026

How Grok powers Vapi's flagship voices

Grok is the default voice engine for Vapi's 12 core voices, across more than 2.5 million agents.

Vapi runs more than 2.5 million voice agents. Grok is now the default voice engine for Vapi's 12 core voices.

0.0M+
voice agents
0
flagship custom voices
#0
in Vapi's blind evaluation

Frontier humanness

Vapi ran an independent, blind evaluation of Grok Voice against other providers. Grok took the #1 spot in that head-to-head arena.

The same model family held up outside Vapi. In a side-by-side poll on X, more than 4,500 people tried to tell a Grok clone from the human original. They split 50/50.

What this means for Vapi developers

If you're already building on Vapi, Grok is the default for the 12 core voices — select it in the dashboard. In Vapi's arena, those conversations scored as more natural, with more emotional range, than the other providers in the eval.

Grok Speech-to-Text and Text-to-Speech are both in the Vapi dashboard, on the same path as the default voices.

More customer stories

Gopuff logo on an orange gradient

How Gopuff's Go agent shops with Grok

LiveKit logo on a green gradient

How Grok Voice powers LiveKit Agents

X logo on a yellow gradient

How X captions 310,000 videos a day with Grok Speech to Text