Scribe v3
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 3
- ElevenLabs Scribe v3 ranks second in non-streaming speech recognition
ElevenLabs launched Scribe v3, which recorded a 1.95% word error rate on Artificial Anlys' AA-WER Index for non-streaming speech-to-text. The result placed it second behind Alibaba's Fun-Realtime-ASR-preview at 1.73% and just ahead of Microsoft AI's MAI-Transcribe-2 at 2.04%, while improving on Scribe v2's 2.18% WER.
ArtificialAnlys (X) · 🔥 4 - ElevenLabs launches Scribe v3 speech-to-text models
ElevenLabs released Scribe v3 for prerecorded audio and Scribe v3 Live for low-latency real-time transcription, with both models available through ElevenAPI and ElevenAgents. Artificial Analysis ranked Scribe v3 among the most accurate transcription models but found it slower and more expensive than several rivals; its benchmark did not test real-time performance, speaker diarization, or rare-language accuracy.
The Decoder · 🔥 0 - ElevenLabs Launches Scribe v3 Speech-to-Text Models
ElevenLabs introduced Scribe v3 and Scribe v3 Live, calling them its most accurate speech-to-text models to date. The company says they are designed for extreme background noise, sentence-ending pauses, and busy multi-speaker meetings. Another supplied report says the models produce 27% fewer errors in more than 90 languages, but no independent validation, baseline, error metric, test set, or methodology is provided.
ElevenLabs (X) · 🔥 0
Experience and discussion from the community
Share my Scribe v3 experienceAsk about Scribe v3
Nobody has shared their experience with Scribe v3 yet.