Submit ProductSign in
Speech to text for recorded and realtime audio
Grounded in available product and source data
Real world recordings are messy. People talk over each other, callers pause mid sentence and a conversation may switch language halfway through. ElevenLabs built Scribe v3 for recorded audio and Scribe v3 Live for realtime audio around those conditions.
Both models are reachable through ElevenAgents and ElevenAPI, where moving from Scribe v2 is a one line change to the model ID. You can set key terms, control whether numbers come back as digits or words, and add formatting rules in plain English. Live adds diarization and a multilingual mode for conversations that change language mid sentence.
