Scribe v3 by ElevenLabs

Name: Scribe v3 by ElevenLabs
Author: ElevenLabs

Scribe v3 and Scribe v3 Live are ElevenLabs' new speech-to-text models: v3 handles recorded audio in 102 languages and v3 Live is the low-latency version that adds speaker labels and context-aware turn detection to realtime transcription.

Visit Scribe v3 by ElevenLabs ↗

Paid (usage-based)ProprietaryREST API / SDKsby ElevenLabs

Why it matters

Teams building meeting notes, call centers, clinical notes and voice agents depend on accurate transcripts and speaker labels. Scribe v3 targets crowded meetings, mid-sentence pauses and code-switched speech, and upgrading is a model-ID change.

Highlights

ElevenLabs reports 27% fewer word errors than Scribe v2 across 102 languages, and 43% fewer errors for v3 Live versus Scribe v2 Realtime.
Speaker diarization comes to realtime transcription for the first time; ElevenLabs reports 17% fewer diarization errors than v2 on recordings with three or more speakers.
v3 Live turn detection uses conversation context as well as silence, so a caller pausing mid-number isn't cut off.
Keyterm prompting takes up to 1,000 terms, transcript formatting rules can be written in plain English, and v3 Live adds a multilingual mode and a larger keyterm library.
Available in ElevenAPI as scribe_v3 and scribe_v3_live, and v3 Live is in ElevenAgents; EU, India and Singapore data residency and Zero Retention Mode for enterprise.

Taken from the project's own README, docs or launch page. Claims about speed or accuracy are the maker's unless stated otherwise.

Try it if: Your app transcribes meetings or phone calls and speaker labels or cut-off callers are a problem.

Worth knowing: All accuracy figures are ElevenLabs' own benchmarks, partly on an internal dataset not yet published.

Quick facts

Pricing	API billed by usage (Scribe v3 rates not checked)not verified
Free or paid	Paid (usage-based)
License	Proprietary
Language	REST API / SDKs
Platform	ElevenLabs app, ElevenAPI and ElevenAgents
GitHub stars	n/a (not a GitHub project)
Maker	ElevenLabs

Related ranking

See how established tools compare in Best Voice & Text-to-Speech. New releases aren't scored until we review them against the rubric.

In the news

ElevenLabs launches Scribe v3 and Scribe v3 Live speech-to-text models with realtime speaker labels November 10, 2026

What we couldn't verify

quick_facts.pricing (per-minute rates for v3)

Checked 2026-11-11 from the model's own pages. Prices and star counts change, so confirm on the official site. Listing here is not a ranking or an endorsement.