ElevenLabs launches Scribe v3 and Scribe v3 Live speech-to-text models with realtime speaker labels
ElevenLabs released Scribe v3 for recorded audio and Scribe v3 Live for realtime audio on November 10, 2026, adding speaker diarization to live transcription and context-aware turn detection for voice agents.
What happened
- ElevenLabs reports Scribe v3 makes 27% fewer word errors than Scribe v2 across 102 languages, and Scribe v3 Live makes 43% fewer errors than Scribe v2 Realtime.
- Scribe v3 Live labels who said what in realtime for the first time and uses conversation context, not just silence, to decide when a caller has finished speaking.
- Keyterm prompting accepts up to 1,000 terms, and formatting rules can be written in plain English; v3 Live adds a multilingual mode for speakers who switch languages mid-sentence.
- Both models are in ElevenAPI as scribe_v3 and scribe_v3_live, v3 Live is available in ElevenAgents, and both are the default in the ElevenLabs app.
Why it matters: Businesses running meeting-notes tools, call centers or AI receptionists depend on transcripts that get names, numbers and speakers right. Upgrading is a model-ID change, though the accuracy figures are ElevenLabs' own benchmarks.
Sources
- ElevenLabs: Introducing Scribe v3 and v3 Live elevenlabs.io
Checked November 11, 2026 against the sources above. Claims about performance or benchmarks are the company's own unless stated otherwise.