Grok Voice ranks #1 on Artificial Analysis’ Task Success Rate leaderboard
Grok Voice Think Fast 2.0 (High) scored 94.6%
Outperforming GPT-Live/Astra, Gemini Live, GPT-Realtime, and every other model on the leaderboard
And this benchmark actually measures whether the voice agent successfully completes the task with the correct final tool call
Not just how good the voice sounds...whether it actually gets the job done
Grok is leading at #1
Grok Voice Think Fast 2.0 (High) scored 94.6%
Outperforming GPT-Live/Astra, Gemini Live, GPT-Realtime, and every other model on the leaderboard
And this benchmark actually measures whether the voice agent successfully completes the task with the correct final tool call
Not just how good the voice sounds...whether it actually gets the job done
Grok is leading at #1
