- Published on
SpaceXAI Releases Grok Voice Transcribe 2.0: A Speech-to-Text API Claiming 2x Accuracy Over 1.0 at $0.10 per Hour
- Authors
- Name
- DeepHorus
SpaceXAI has launched Grok Voice Transcribe 2.0, a speech-to-text API it says is twice as accurate as version 1.0 at the same price. The model targets hard audio: noisy phone lines, competing voices, local accents, and spoken credentials. It is live today as a hosted API under the model ID grok-voice-transcribe-2.0, with no open weights.
The model is built on the audio foundation behind Grok Voice, which handles tens of thousands of customer-support calls a day. It also transcribes millions of hours of video narration and powers the Grok assistant in Tesla vehicles. Training data came from live, noisy, multilingual audio, followed by post-training refinement.
SpaceXAI reports a first-place accuracy rank among 32 streaming models on the public Artificial Analysis leaderboard. On internal production sets, version 2.0 improves on 1.0 across telephony, conversational speech, credentials, and short phrases. Short-phrase word error rate drops from 20.6% to 6.8%, about 67% fewer errors.
The API supports batch and streaming, word-level timestamps, speaker diarization, and up to eight channels. It transcribes dozens of languages, detects language automatically, and follows mid-recording switches. Batch files up to 500 MB are accepted across 12 audio formats.
Pricing is unchanged: $0.10 per hour for batch and $0.20 per hour for streaming. Diarization, timestamps, and key terms are included. That equals about $1.67 and $3.33 per 1,000 minutes.
Atlassian Loom has adopted the model to transcribe every video, finding it more accurate than its existing solution. The workflow is record, transcribe, then code. However, SpaceXAI's benchmarks are vendor-reported and not independently reproduced.
The release gives developers a low-cost option for multilingual speech-to-text, especially for voice agents and in-car commands. Self-hosting is not available, so users must rely on the hosted API. The model is live today, positioning SpaceXAI to compete in the growing STT market.
