ElevenLabs launches Scribe v3 with 27% fewer word errors than v2 across 102 languages
iamfakhrealam · x · 2026-11-14
ElevenLabs released two new speech-to-text models:
- Scribe v3, for recorded audio, makes 27% fewer word errors than Scribe v2 across 102 languages, with the most accurate speaker labels the team tested
- Scribe v3 Live, the low-latency realtime variant, cuts errors by 43% vs Scribe v2 Realtime and adds speaker diarization plus improved turn detection
Independent evaluator voicearena tested on 70+ hours of real, unscripted phone calls and found Scribe v3 Live takes #1 in Romanian and Indian English streaming, topping all 10 metric splits it reports. The key point: vendor benchmarks use read speech, but spontaneous conversation is where models struggle most.
Scribe v3 Live is live in ElevenAgents; both models are available via ElevenAPI with a one-line model ID change from v2.
More from Models
- Claude iOS hit by Sign in with Apple outage affecting new signups — ClaudeAI-mod-bot · 2026-11-14
- Independent scan: 185 of 265 open LLMs can infer their training GPU family — willdepue · 2026-11-14
- How credible are frontier labs' claims of solving Navier–Stokes and Fermat's Last Theorem? — Witty-Ad-8579 · 2026-11-14
- Users pile on OpenAI: buggy desktop app and DevDay flop push them to Claude — koltregaskes · 2026-11-14
- Heavy AI users stuck between Pro and Ultra tiers as daily usage lands mid-tier — gaganghotra_ · 2026-11-14
- Google Search reportedly rolling out Gemini 4 Hydrogen globally with AI-first results — feblet · 2026-11-14