xAI Launches Grok Voice Transcribe 2.0 With Doubled Accuracy
Source xAI News (mirror)
At a glance
xAI has released Grok Voice Transcribe 2.0, a new model that converts speech into text. The company says it is twice as accurate as the previous version while keeping the same price. It was designed to understand speech well even in noisy sounds such as phone calls.
Key points
The company says it is twice as accurate as the previous version, 1.0, in real-world tests.
It automatically detects multiple languages and keeps up when the language changes in the middle of speech.
Batch processing costs $0.10 per hour of audio, and real-time processing costs $0.20 per hour, at the same price as before.
Services using the existing speech-to-text API can get the accuracy improvements without changing their code. Version 1.0 will be discontinued within a few weeks.
Why it matters
More and more meetings and customer calls are being transcribed, so more accurate transcription makes it easier to use in real work.
Original article
Introducing Grok Voice Transcribe 2.0
xAI News (mirror) · Sep 18, 2026, 9:00 AM GMT+9 · x.ai
Read on xAI News (mirror)AI read the original article and summarized it in plain language. It may contain mistakes, so please check the original for accuracy. The article's copyright belongs to xAI News (mirror).