xAI / SpaceX Super IntelligenceGrok Voice
Grok Voice Transcribe 2.0
The current transcription model supports more than 38 languages, file uploads and real-time WebSocket audio.
What the release provides
The API offers word timestamps, multiple audio channels and optional text formatting. Language detection can follow language changes. A supplied language code guides ambiguous audio; formatting support covers a smaller language list.Grok speech-to-text API and model selection
How to use it well
Practical guidance. Preserve the raw transcript when formatting numbers, dates or currencies. A plausible formatted value can be different from what the speaker said. Review low-quality audio and names against the recording before using the transcript as a formal record.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | grok-voice-transcribe-2.0 |