Skip to main content

ElevenLabsAudio modelActive

Scribe v2

PopularityRank 283 of 876

39out of 100same as last month

Summary

Scribe v2 is ElevenLabs’ asynchronous speech recognition model for turning recordings into usable transcripts. It covers more than 90 languages and adds word-level timestamps, speaker diarization for up to 32 speakers, and entity detection. Keyterm prompts accept up to 1,000 names or domain terms and add a controlled vocabulary to the transcription.

Scribe v2 returns timestamps, speaker labels, audio tags, and transcript editing alongside the raw text. The model handles completed or uploaded audio files; Scribe v2 Realtime is the separate endpoint for a live audio stream.

GraduallyBench Audio

Speech recognition compared automatically.

The pilot uses the same 32 read-speech recordings in German and English, plus four digital-silence clips. The comparison uses source transcripts without human verification. Each language contains four speakers. Spontaneous speech and background noise are outside this cohort.

Word error rate (WER) counts replaced, omitted and inserted words relative to the reference words. Character error rate (CER) compares characters. Lower values mean fewer differences from the reference. WER can exceed 100 %. Counts below each rate show errors / reference units. Silence is evaluated separately.

Protocol audio-stt-read-speech-automated-v1. Actual provider billing is unknown.

Scribe v2

fal.ai · fal-ai/elevenlabs/speech-to-text/scribe-v2 · Tested Nov 5, 2026

Successful requests 36 / 36. The exact backend version is not confirmed by the requested route.

Read-speech transcription
LanguageSuccessful responsesWERCERWER, speaker mean
German16 / 163.64 %20 / 5500.98 %32 / 32623.27 %n = 4 / 4
English16 / 161.63 %8 / 4900.62 %16 / 25801.68 %n = 4 / 4
Digital silence
Language settingClips with invented speechInvented words
DE0 / 20
EN0 / 20

Request duration, median / p95 2,745 / 6,215 ms (n = 36). Includes network time and client processing.

All benchmark results

Every published result with its source, test setup and comparison models.

Show all 84 published benchmark results

Page 1 of 7

Measurements without matching peer values

These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.

Artificial Analysis AA-WER (non-streaming) (Show measurement, test conditions, and source)
Source value
2.2
Score
2.2
Metric
word error rate (WER)
Unit
%
Benchmark version
AA-WER v2.2
Category
audio
Direction
Lower is better
Source type
independent-evaluator
Evaluator
Artificial Analysis
Status
active
Retrieved at
2026-10-15
Methodology
Independent evaluation. Artificial Analysis directly tested the named API endpoint; audio-duration-weighted WER across AA-AgentTalk (50%), VoxPopuli-Cleaned-AA English (25%), and Earnings22-Cleaned-AA (25%), about eight hours total. Current normalizer AA-WER v2.2.
Source Artificial Analysis
ElevenLabs MedDictate overall WER (Show measurement, test conditions, and source)
Source value
7.6
Score
7.6
Metric
overall WER
Unit
%
Benchmark version
Corti MedDictate; evaluator edition not stated
Category
audio
Direction
Lower is better
Sample
40
Source type
vendor-reported
Evaluator
ElevenLabs model evaluation
Status
active
Retrieved at
2026-10-15
Methodology
Vendor-reported result. Vendor evaluation on Corti MedDictate: English, French, and German clinical dictation. The public test split has 40 clips (24 EN, 7 FR, 9 DE); overall aggregation details are not stated.
Source ElevenLabs model evaluation
ElevenLabs MedTerm medical-term recall (Show measurement, test conditions, and source)
Source value
77.1
Score
77.1
Metric
recall of annotated medical terms
Unit
%
Benchmark version
Corti MedTerm (600 samples)
Category
audio
Direction
Higher is better
Sample
600
Source type
vendor-reported
Evaluator
ElevenLabs model evaluation
Status
active
Retrieved at
2026-10-15
Methodology
Vendor-reported result. Vendor evaluation on Corti MedTerm, a public 600-sample clinical-terminology benchmark; term-WER scores annotated medical terms only.
Source ElevenLabs model evaluation
ElevenLabs MedTerm medical-term WER (Show measurement, test conditions, and source)
Source value
10.7
Score
10.7
Metric
term-WER
Unit
%
Benchmark version
Corti MedTerm (600 samples)
Category
audio
Direction
Lower is better
Sample
600
Source type
vendor-reported
Evaluator
ElevenLabs model evaluation
Status
active
Retrieved at
2026-10-15
Methodology
Vendor-reported result. Vendor evaluation on Corti MedTerm, a public 600-sample clinical-terminology benchmark; term-WER scores annotated medical terms only.
Source ElevenLabs model evaluation
ElevenLabs Eka Medical ASR semantic WER (Show measurement, test conditions, and source)
Source value
7.35
Score
7.35
Metric
semantic WER
Unit
%
Benchmark version
Eka Medical ASR; evaluator edition not stated
Category
audio
Direction
Lower is better
Sample
3619
Source type
vendor-reported
Evaluator
ElevenLabs model evaluation
Status
active
Retrieved at
2026-10-15
Methodology
Vendor-reported result. Vendor evaluation on public Eka Medical ASR, 3,619 English clinical-audio samples; semantic WER counts meaning-equivalent formatting as equivalent.
Source ElevenLabs model evaluation
ElevenLabs Eka Medical ASR keyword WER (Show measurement, test conditions, and source)
Source value
7.04
Score
7.04
Metric
keyword WER
Unit
%
Benchmark version
Eka Medical ASR; evaluator edition not stated
Category
audio
Direction
Lower is better
Sample
3619
Source type
vendor-reported
Evaluator
ElevenLabs model evaluation
Status
active
Retrieved at
2026-10-15
Methodology
Vendor-reported result. Vendor evaluation on public Eka Medical ASR, 3,619 English clinical-audio samples; keyword-WER scores annotated medical keywords only.
Source ElevenLabs model evaluation
Appen MedTerm-90 overall clinical entity recall (Show measurement, test conditions, and source)
Source value
83.35
Score
83.35
Metric
Recall of clinician-validated medical entities
Unit
%
Benchmark version
MedTerm-90, September 2026 model versions
Category
audio
Direction
Higher is better
Sample
30017
Lower confidence bound
82.6
Upper confidence bound
84.09
Source type
independent-evaluator
Evaluator
Appen MedTerm-90
Status
active
Retrieved at
2026-10-16
Methodology
Independent evaluation. Appen reports the September 2026 ElevenLabs Scribe v2 result on 1413 real physician dictations spanning 90.5 hours and 14 specialties. The reported recall is over 30017 clinician-validated entities; this entity total is the reported count, not the number of recordings. Identical files, entities, thresholds, and judges were used for all models, without vocabulary hints. Appen reports a bootstrap 95% interval.
Context
This is clinical-entity recall, not WER or a general transcription score. Results represent September 2026 model versions under close-microphone single-speaker dictation conditions; Appen notes these conditions favor the models and may not generalize to other audio. The report page gives only an overall score, not the advertised per-specialty breakdown.
Source Appen MedTerm-90
Superwhisper speech benchmark macro word error rate (Show measurement, test conditions, and source)
Source value
7.5
Score
7.5
Metric
macro word error rate (WER) over eight speech datasets
Unit
%
Benchmark version
Superwhisper speech benchmark, model runs 20260731-174525 / 20260811 dataset runs; page captured 2026-10-17
Category
speech-to-text / batch transcription
Direction
Lower is better
Source type
independent-evaluator
Evaluator
Superwhisper speech benchmark
Status
active
Retrieved at
2026-10-17
Methodology
Independent evaluation. Exact product ElevenLabs Scribe v2 (hosted batch API); hosted batch speech-to-text; evaluator runs the same audio and machines as its other model rows; reported score is macro-average word error rate over 8 named datasets. No medical or realtime route is implied.
Context
The page publishes the aggregate and the eight-dataset count, not a clip-level denominator, per-dataset weighting details, confidence interval or immutable serving revision. The run identifiers are preserved as published. This batch/API result is separate from realtime and medical Scribe variants.
Source Superwhisper speech benchmark
Ocular Converse-STT Version 2 pooled WER (Show measurement, test conditions, and source)
Source value
1.22
Score
1.22
Metric
pooled word error rate across the eight speaker tracks
Unit
%
Benchmark version
Converse-STT Version 2; report dated 2026-10-13
Category
speech-to-text
Direction
Lower is better
Source type
independent-evaluator
Evaluator
Ocular AI Converse-STT Version 2
Status
active
Retrieved at
2026-10-17
Methodology
Independent evaluation. Ocular's exact route table identifies Scribe v2 as the ElevenLabs OpenRouter route elevenlabs/scribe-v2 with MP3 input and no non-default parameters. The task is batch, single-channel, single-speaker transcription of the full 15-minute tracks, without mixing, diarization or a streaming requirement. The result pools edit counts across four two-person American English conversations, eight speaker tracks and 12,555 reference words, with Whisper's English normalizer whisper-normalizer 0.1.12.
Context
This is a small, fixed public sample, not a population estimate. The 12,555-word denominator is not treated as sampleCount. The API-serving revision is not pinned.
Source Ocular AI Converse-STT Version 2
Omi PriMock57 mean per-file WER (Show measurement, test conditions, and source)
Source value
9.720442500254512
Score
9.720442500254512
Metric
Arithmetic mean of per-file word error rates
Unit
%
Benchmark version
Omi Health PriMock57, pinned commit c0e6c11528ec150f5ff98627f8e86c2d6e731c33, README updated 2026-04-29
Category
speech-to-text
Direction
Lower is better
Sample
55
Source type
independent-evaluator
Evaluator
Omi Health medical-STT-eval
Status
active
Retrieved at
2026-10-17
Methodology
Independent evaluation. Exact field model_name is elevenlabs-scribe_v2; 55 PriMock57 files and 80,709 reference words. Custom Whisper-based EnglishTextNormalizer with oh-fix and post-word-map. This is average per-file WER, not pooled corpus WER.
Context
Omi labels the tested endpoint elevenlabs-scribe_v2, the standard Scribe v2 API, not the separate scribe_v2_medical model_id. The repository is pinned, but the API serving revision is not. No confidence interval is published. PriMock57 contains medical dialogue and is not a general-domain corpus.
Source Omi Health medical-STT-eval
Omi PriMock57 global medical M-WER (Show measurement, test conditions, and source)
Source value
2.5397
Score
2.5397
Metric
Canonical Medical WER v2 over medical reference tokens
Unit
%
Benchmark version
Omi Health PriMock57, pinned commit c0e6c11528ec150f5ff98627f8e86c2d6e731c33, README updated 2026-04-29
Category
speech-to-text
Direction
Lower is better
Sample
55
Source type
independent-evaluator
Evaluator
Omi Health medical-STT-eval
Status
active
Retrieved at
2026-10-17
Methodology
Independent evaluation. Standard Scribe v2 API model elevenlabs-scribe_v2, not medical endpoint; 55 files and 3,150 medical reference tokens. Canonical M-WER v2 uses narrowly reviewed reference corrections and morphology while penalizing clinical confusables.
Context
Omi labels the tested endpoint elevenlabs-scribe_v2, the standard Scribe v2 API, not the separate scribe_v2_medical model_id. The repository is pinned, but the API serving revision is not. No confidence interval is published. PriMock57 contains medical dialogue and is not a general-domain corpus.
Source Omi Health medical-STT-eval
Omi PriMock57 global drug M-WER (Show measurement, test conditions, and source)
Source value
2.7897
Score
2.7897
Metric
Canonical Medical WER v2 on drug reference tokens
Unit
%
Benchmark version
Omi Health PriMock57, pinned commit c0e6c11528ec150f5ff98627f8e86c2d6e731c33, README updated 2026-04-29
Category
speech-to-text
Direction
Lower is better
Sample
55
Source type
independent-evaluator
Evaluator
Omi Health medical-STT-eval
Status
active
Retrieved at
2026-10-17
Methodology
Independent evaluation. Standard Scribe v2 API model elevenlabs-scribe_v2; 55 files and 466 drug reference tokens; Canonical M-WER v2 retains high-risk drug confusables as errors.
Context
Omi labels the tested endpoint elevenlabs-scribe_v2, the standard Scribe v2 API, not the separate scribe_v2_medical model_id. The repository is pinned, but the API serving revision is not. No confidence interval is published. PriMock57 contains medical dialogue and is not a general-domain corpus.
Source Omi Health medical-STT-eval

Page 1 of 7

Popularity compared

How often the model is searched for, used and mentioned, compared with every model of its category.

Popularity index

0 to 100, higher is more popular

20 of 83 models

pyannote Community-173
Nemotron-3-Diarization57
Gemini 3.5 Transcribe52
Confucius4-R2T252
NVIDIA Nemotron 3.5 Transcribe52
Granite Speech 5.0 470M TurboCTC52
Kokoro-82M51
Nova-351
Fish Audio S2 Pro51
Granite Speech 5.0 470M TurboCTC NC50
Orukeet49
Lyria 3.547
Gemini 3.5 Live Translate46
Phonon-244
Aura-244
Gemini 3.1 Flash TTS44
MiniMax-Music341
Audio8-ASR-Infinite41
GPT-Realtime 2.141
Scribe v239

As of November 2026 · Source: gradually.ai popularity index

Over time

Weekly values from October 18, 2026 to November 6, 2026

Scribe v2Maya 2 Native 40Eleven v3 39GPT-Live-Transcribe 39Sonic 3.6 39

Method changes in the timeline, a jump at these points is no market event:

  1. 1November 1, 2026: Value only with search or usage data, 30-day attention

What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.

Scribe v2 compared

Similar models

More models of the same kind as Scribe v2, chosen by model line, developer, size, price and release date.

All specifications

Architecture

Model class
Speech recognition
Source: Model class

Input and output

Input
Audio
Source: Input
Output
Text
Source: Output

Watermarking and provenance

AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.

AI watermarking
Unknown
no source
Watermarking technique
Not known
no source

Licence and access

Access
API
Source: Access

API

API model ID
scribe_v2
Source: API model ID

Further details

Task
Transcription
Source: Task
Supported details
Asynchronous transcription
Source: Supported details
Transcription
Over 90 languages, word timestamps and up to 32 speakers
Source: Transcription
Keyterms
Up to 1,000 terms as keyterm prompts
Source: Keyterms
Entities
65 entity types
Source: Entities

Sources and data date

Show all 41 sources

Benchmark sources at a glance

Which independent evaluators tested this model. This is coverage, not a quality ranking.

EvaluatorTest seriesComposite indicesRecorded active measurements
Artificial Analysis001

Without vendor claims and historical or disputed results; those stay labelled in the single values. 1 measurements are not yet assigned to a test series.

TypeEvidence and data date
Research statusResearch date October 15, 2026. 36 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not published

clip-level-denominator-and-serving-revision

Note

Clip-level denominator and immutable serving revision are not published.

Not published

independent-conversation-sample-and-serving-revision

Note

Only four conversations were evaluated; a population-representative sample and an immutable serving revision are not available.

Additional sourceElevenLabs model documentation (retrieved September 24, 2026; October 15, 2026; October 17, 2026) · ElevenLabs Models documentation · Editorial description reviewed October 2, 2026
Additional sourceArtificial Analysis (retrieved October 15, 2026)
Additional sourceElevenLabs model evaluation (retrieved October 15, 2026; October 17, 2026; October 18, 2026)
Additional sourceAppen MedTerm-90 (retrieved October 16, 2026)
Additional sourceSuperwhisper speech benchmark (retrieved October 17, 2026)
Additional sourceOcular AI Converse-STT Version 2 (retrieved October 17, 2026; October 22, 2026)
Additional sourceOmi Health medical-STT-eval (retrieved October 17, 2026)123
Additional sourceSierra μ-bench (retrieved October 17, 2026)
Additional sourceSierra Research, mu-bench (retrieved October 17, 2026)
Additional sourceBenchmarking Commercial ASR Systems on Code-Switching Speech (retrieved October 17, 2026)
Additional sourceOmi Health Medical Speech-to-Text Benchmark (retrieved October 17, 2026; October 18, 2026)
Additional sourceAssemblyAI speech benchmark (retrieved October 18, 2026)
Additional sourceNyra Verbatim Speech Benchmark (retrieved October 18, 2026)
Additional sourceDeepgram Nova-3 Medical Multilingual article (retrieved October 18, 2026)
Additional sourceOpenbenchmarks (retrieved October 21, 2026)
Additional sourceFA-Bench (Olewave) (retrieved October 22, 2026)
Additional sourceHugging Face Open ASR Leaderboard (retrieved October 28, 2026)
Additional sourceModel metadata (retrieved October 15, 2026)12345
Additional sourceModel metadata (retrieved October 17, 2026)123456789
Additional sourceModel metadata (retrieved October 28, 2026)12
Additional sourceOpenRouter: Popularity index (derived), CC BY 4.0
Additional sourceVercel AI Gateway: Popularity index (derived), CC BY 4.0
Additional sourceOllama: cumulative downloads for the popularity index
Additional sourceReplicate: cumulative runs for the popularity index
Additional sourceGoogle Trending Now: limited feed selection for Germany and the US
Additional sourceCreative Commons Attribution 4.0