Tech Meridian ← LIVE FEED
PROMY MERIDIAN RU

RELEASE · MODELS · #3477

ElevenLabs releases Scribe v3 and Scribe v3 Live speech-recognition models

ElevenLabs launched Scribe v3 for pre-recorded audio and Scribe v3 Live for low-latency real-time use; both target noisy, overlapping speech and include improved speaker diarization and multilingual handling. Independent benchmarks from Artificial Analysis place Scribe v3 among the most accurate transcription models, though it is slower and pricier than some competitors, and ElevenLabs is expanding enterprise ties with Nvidia and McKinsey.

KEY POINTS

  1. ElevenLabs launched Scribe v3 for pre-recorded audio and Scribe v3 Live for low-latency real-time use; both target noisy, overlapping speech and include improved speaker diarization and multilingual handling.
  2. Independent benchmarks from Artificial Analysis place Scribe v3 among the most accurate transcription models, though it is slower and pricier than some competitors, and ElevenLabs is expanding enterprise ties with Nvidia and McKinsey.
  3. Real-time conversational-aware transcription (Scribe v3 Live) can reduce voice agents cutting off speakers and improve reliability in noisy, multi-speaker settings, affecting enterprise voice products and deployments.

WHY IT MATTERS

Real-time conversational-aware transcription (Scribe v3 Live) can reduce voice agents cutting off speakers and improve reliability in noisy, multi-speaker settings, affecting enterprise voice products and deployments.

SOURCES & TIMELINE

1