Audio modelPublic preview

Voxtral Mini Transcribe Realtime 3

Mistral AI

Released: October 20, 2026
Data date: October 24, 2026

Access: Mistral API; real-time mode over WebSocket

Voxtral Mini Transcribe Realtime 3 streams transcripts over a WebSocket connection. Mistral documents configurable audio chunks from 80 to 2,400 milliseconds, speaker labels, and end-of-turn detection.

Specifications and access

Specification	Value and source
API model ID	voxtral-mini-transcribe-realtime-3-0Source
Release	Listed in the audio overview; no separate announcement or release date specifiedSource
Access	Mistral API; real-time mode over WebSocketSource
Input	Streaming audioSource
Output	Streaming transcript with speaker labels and end-of-turn detectionSource
Chunk latency	Configurable from 80 to 2,400 millisecondsSource
Release	2026-10-20Source
Audio chunk latency in milliseconds	80-2400Source
Price per audio minute in USD	0.008Source
Cached price per audio minute in USD	0.0008Source
Omi report edition	Public data identifies edition 2026-10-18.1, 1,513 English recordings, 7.2096 audio hours and suite 2.0.0.Source
Model ID in the Omi evaluation	Internal Omi profile key voxtral-mini-transcribe-realtime-3. It is not an established API model or checkpoint identifier. (The page does not identify this as the exact Mistral API model ID and exposes no request metadata.)Source
Omi reference corpus	1,513 recordings; 7.2096 hours; 63,815 reference words; 2,872 medical-term references; 442 drug-name references; 94 dosage/unit reference events.Source
Omi model row date	2026-10 accuracy; realtime timing 2026-10-19. Accuracy is dated only to the month; the timing run has a separate day-level date.Source
Language support in the current documentation	The language table lists voxtral-mini-transcribe-realtime-3-0 with support for 45 languages and dialects.Source
Model ID in the API examples	Official request docs show the model string voxtral-mini-transcribe-realtime-3. (The model card uses voxtral-mini-transcribe-realtime-3-0; the request documentation omits trailing '-0' and does not explain whether they are aliases.)Source
Documented realtime interface	WebSocket through /v1/audio/transcriptions/realtime, with the realtime model passed as a query parameter.Source

Benchmark sources at a glance

Each evaluator measures different capabilities under its own conditions. Test series, composite indices and their components stay separate. The detailed results below retain their scales, configurations and source dates.

No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.

Coverage is not a quality ranking. Overlapping tests and multiple settings do not count as independent experiments. Vendor claims, historical and disputed results are excluded from this overview and remain labelled in the detailed evidence.

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.

Omi Medical realtime: Medical-term WER (Show measurement, test conditions, and source)

Source value: 1.36
Score: 1.36
Metric: Medical-term WER
Unit: %
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Lower is better
Sample: 2872
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime transcription on the 1,513-recording English medical corpus. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. Metric denominator: 2872.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

Omi Medical realtime: Overall WER (Show measurement, test conditions, and source)

Source value: 7.44
Score: 7.44
Metric: Overall WER
Unit: %
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Lower is better
Sample: 63815
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime transcription on the 1,513-recording English medical corpus. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. Metric denominator: 63815.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

Omi Medical realtime: Drug-name errors (Show measurement, test conditions, and source)

Source value: 1.36
Score: 1.36
Metric: Drug-name errors
Unit: %
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Lower is better
Sample: 442
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime transcription on the 1,513-recording English medical corpus. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. Metric denominator: 442.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

Omi Medical realtime: Dose/unit matches (Show measurement, test conditions, and source)

Source value: 75.53
Score: 75.53
Metric: Dose/unit matches
Unit: %
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Higher is better
Sample: 94
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime transcription on the 1,513-recording English medical corpus. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. Metric denominator: 94.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

Omi Medical realtime: Dose/unit errors (Show measurement, test conditions, and source)

Source value: 24.47
Score: 24.47
Metric: Dose/unit errors
Unit: %
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Lower is better
Sample: 94
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime transcription on the 1,513-recording English medical corpus. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. Metric denominator: 94.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

Omi Medical realtime: Dose precision (Show measurement, test conditions, and source)

Source value: 92.21
Score: 92.21
Metric: Dose precision
Unit: %
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Higher is better
Sample: 77
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime transcription on the 1,513-recording English medical corpus. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. Metric denominator: 77. Precision denominator is the number of predicted dose/unit events.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

Omi Medical realtime: Dose F1 (Show measurement, test conditions, and source)

Source value: 83.04
Score: 83.04
Metric: Dose F1
Unit: %
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Higher is better
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime transcription on the 1,513-recording English medical corpus. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. F1 combines reference and predicted events.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

FLEURS 35 realtime WER at 2400 ms chunk latency (Show measurement, test conditions, and source)

Source value: 5.7
Score: 5.7
Metric: WER
Unit: %
Category: realtime speech recognition
Direction: Lower is better
Source type: vendor-reported
Evaluator: Mistral Voxtral 3 announcement and benchmark tables
Status: active
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Realtime transcription model; FLEURS 35 macro-average, streaming chunk latency set to 2400 ms. The chart's remaining models are realtime.

Source Mistral Voxtral 3 announcement and benchmark tables

FLEURS 35 realtime WER at 960 ms chunk latency (Show measurement, test conditions, and source)

Source value: 6
Score: 6
Metric: WER
Unit: %
Category: realtime speech recognition
Direction: Lower is better
Source type: vendor-reported
Evaluator: Mistral Voxtral 3 announcement and benchmark tables
Status: active
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Realtime transcription model; FLEURS 35 macro-average, streaming chunk latency set to 960 ms. The chart's remaining models are realtime.

Source Mistral Voxtral 3 announcement and benchmark tables

FLEURS 35 realtime WER at 480 ms chunk latency (Show measurement, test conditions, and source)

Source value: 6.4
Score: 6.4
Metric: WER
Unit: %
Category: realtime speech recognition
Direction: Lower is better
Source type: vendor-reported
Evaluator: Mistral Voxtral 3 announcement and benchmark tables
Status: active
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Realtime transcription model; FLEURS 35 macro-average, streaming chunk latency set to 480 ms. The chart's remaining models are realtime.

Source Mistral Voxtral 3 announcement and benchmark tables

FLEURS 35 realtime WER at 240 ms chunk latency (Show measurement, test conditions, and source)

Source value: 7.4
Score: 7.4
Metric: WER
Unit: %
Category: realtime speech recognition
Direction: Lower is better
Source type: vendor-reported
Evaluator: Mistral Voxtral 3 announcement and benchmark tables
Status: active
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Realtime transcription model; FLEURS 35 macro-average, streaming chunk latency set to 240 ms. The chart's remaining models are realtime.

Source Mistral Voxtral 3 announcement and benchmark tables

Omi Medical realtime: Median final latency (Show measurement, test conditions, and source)

Source value: 396
Score: 396
Metric: Median final latency
Unit: ms
Benchmark version: Omi report edition 2026-10-18.1
Category: medical-transcription
Direction: Lower is better
Sample: 199
Source type: vendor-reported
Evaluator: Omi Medical benchmark v1.1
Status: disputed
Retrieved at: 2026-10-24
Methodology: Vendor-reported result. Rank 5 in Omi's realtime table. Realtime from shared end-of-speech to final transcript; 200 short-clip attempts, one invalid attempt excluded from timing (N=199); AWS us-east-2, October 19, 2026. Rank 5; Omi internal row profile voxtral-mini-transcribe-realtime-3. Metric denominator: 199.
Context: Omi labels its report edition 2026-10-18.1 and Voxtral accuracy rows only by month (2026-10); batch timing is dated October 18 and realtime timing October 19, before Mistral announced public preview on October 20. The exact endpoint/build is not identified. These are disputed observations, not definitive measurements of the post-release model.

Source Omi Medical benchmark v1.1

Page 1 of 2

Sources and data date

Type	Evidence and data date
Research status	Research date October 24, 2026. 90 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not published	Model weights and license Note The public API model card does not state downloadable weights or an open-source license.
Not found in the inspected sources	Realtime results by language Note The public realtime chart reports FLEURS 35 macro-averages by chunk latency, not per-language scores.
Unresolved	Exact Omi endpoint and model build Note Omi names the product and an internal profile key but not the exact API endpoint or model metadata. The timing conflict prevents a definitive match to the public version announced on October 20.
Unresolved	Model card to request ID crosswalk Note The model card uses voxtral-mini-transcribe-realtime-3-0, while the request example uses voxtral-mini-transcribe-realtime-3. Public documentation does not say whether the identifiers are interchangeable.
Unresolved	Exact independent result for Voxtral Mini Transcribe Realtime 3.0 Note Artificial Analysis publishes a different 480 ms variant ID; no exact API/model bridge to the target profile ID was found. No legacy realtime score is transferred.
Not found in the inspected sources	Open ASR leaderboard result for exact Voxtral Mini Transcribe Realtime 3.0 ID Note The current Open ASR board did not expose this exact ID; older Voxtral 2507 entries are version-mismatched.
Additional source	Mistral audio models overview (retrieved October 24, 2026)
Additional source	Mistral speech transcription overview (retrieved October 24, 2026) · Mistral speech transcription overview · Editorial description reviewed October 24, 2026
Additional source	Mistral model changelog (retrieved October 24, 2026)
Additional source	Mistral inference pricing (retrieved October 24, 2026)
Additional source	Voxtral realtime transcription benchmark charts (retrieved October 24, 2026) · Mistral Voxtral 3 announcement · Editorial description reviewed October 24, 2026
Additional source	Omi Medical benchmark v1.1 (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)
Additional source	Model metadata (retrieved October 24, 2026)

Specifications and access

Benchmark sources at a glance

Measurements without matching peer values

Published prices

Sources and data date