🚀Universal-1 Model Launch

This week, we’ve launched Universal-1, our most powerful and accurate Speech-to-Text model to date, trained on 12.5M hours of multilingual audio data

audio data launch multilingual newsletter speech speech-to-text text universal universal-1