SpaceXAI Releases Grok Voice Transcribe 2.0 with Half the Error Rate at the Same Price

SpaceXAI has released Grok Voice Transcribe 2.0, a speech-to-text model that cuts the error rate by about half while keeping the same price. It is built on the audio foundation model behind Grok Voice.

Grok Voice now handles tens of thousands of customer service calls daily, transcribes millions of hours of video narration, and powers voice agents like the Grok assistant in Tesla vehicles. On the Artificial Analysis leaderboard, the model ranks first in accuracy among all 32 streaming models.

In tests on four internal datasets covering customer calls, daily conversations, spoken details, and short multilingual commands, version 2.0 outperformed 1.0 across the board. Pricing stays the same: $0.10 per hour for batch transcription and $0.20 per hour for real-time streaming, with speaker separation, timestamps, and custom keywords included at no extra cost.