theaicatalog
ElevenLabs ยท Models

ElevenLabs launches Scribe v3 speech-to-text models

ElevenLabs says its new Scribe v3 and Scribe v3 Live speech-to-text models are its most accurate yet. Scribe v3 makes 27% fewer word errors than Scribe v2 across 102 languages on recorded audio, and 17% fewer diarization errors on recordings with three or more speakers. Scribe v3 Live, built for realtime audio, makes 43% fewer errors than Scribe v2 Realtime and adds speaker labeling and turn detection that uses conversation context as well as silence. Developers can upgrade by changing the model ID to scribe_v3 or scribe_v3_live.