AI KNOWLEDGE DESK

Models · entities · concepts · comparisons · practical tools

GETLLMS.ORG

Grok Voice Transcribe 2.0

Grok Voice Transcribe 2.0 is SpaceXAI's current speech-to-text model for batch files and live streaming transcription.

Speech RecognitionAudio TranscriptionStreaming ASRDiarizationMultichannelSpaceXAI
SpaceXAI REST /v1/stt and WebSocket wss://api.x.ai/v1/stt in us-east-1
REST $0.10/hour; streaming $0.20/hour
Commercial

🚀Function Overview

Grok Voice Transcribe 2.0 supports batch and live transcription with timestamps, language formatting, diarization, multichannel audio, keyterms, and Smart Turn.

Key Features

  • Official model ID grok-voice-transcribe-2.0
  • REST file or URL transcription at $0.10 per hour
  • WebSocket streaming transcription at $0.20 per hour
  • 12 documented audio formats and files up to 500 MB
  • Word-level timestamps, detected language, speaker diarization, and up to eight interleaved channels
  • Keyterm biasing, filler-word control, interim results, endpointing, voice-activity threshold, and Smart Turn
  • The live model and capability pages name 2.0 as the default, while the September 17 release note still names 1.0; pin 2.0 explicitly

Use Cases

  • Meeting and interview transcription
  • Call-center multichannel transcription
  • Live captions and dictation
  • Speaker-labelled recordings
  • Domain vocabulary using keyterms

⚙️Input Parameters

model

string

Pass grok-voice-transcribe-2.0 explicitly because first-party default documentation conflicts.

file

file

Audio file for REST transcription; mutually exclusive with url.

url

string

Server-downloadable audio URL for REST transcription; mutually exclusive with file.

language

string

Optional language code that enables inverse text normalization for supported languages.

keyterm

string

Repeatable domain term used to bias transcription.

💡Usage Examples

Example 1

Input Parameters

{
  "model": "grok-voice-transcribe-2.0",
  "file": "meeting.mp3",
  "language": "en",
  "keyterm": "GetLLMs"
}

Output Results

A transcript with detected language, audio duration, and word-level timestamps; speaker or channel fields appear when enabled.

Quick Actions

Technical Specifications

Hardware Type
SpaceXAI REST /v1/stt and WebSocket wss://api.x.ai/v1/stt in us-east-1
Commercial Use
Supported
Pricing
REST $0.10/hour; streaming $0.20/hour