whatmodel.sh / stt / Voxtral Mini Transcribe V2
Voxtral Mini Transcribe V2
↗ share
+ contribute
Voxtral Mini Transcribe V2 is a closed-weights speech-to-text model from Mistral, released February 2026. Licence: proprietary.
Score within the Open ASR English short-form WER (March 2026) comparison set only.
$ how it compares
The speech-to-text models ranked either side of Voxtral Mini Transcribe V2, out of 176 with a score, then more from Mistral.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 44 |  Fun-ASR-Flash | 59±4 | comparison only | ||
| 45 |  Granite Speech 3.3 8B | 59±5 | local | ||
| 46 | Voxtral Mini Transcribe V2 | 59±3 | api $0.0030/audio min | ||
| 47 |  Avalon v1 (en) | 59±5 | comparison only | ||
| 48 |  GPT-Realtime-Whisper | 59±3 | api $0.017/audio min | ||
| more from Mistral | |||||
| 32 | Voxtral Small 24B (2507) | 61±2 | api $0.0040/audio min +local | ||
| 52 | Voxtral Mini 3B (2507) | 58±1 | api $0.0010/audio min +local | ||
| 96 | Voxtral Mini 4B Realtime (2602) | 53±2 | api $0.0060/audio min +local | ||
see all speech-to-text models · find the ones that fit your machine
$ what it's good at
long-form transcription█·····█···██ #2 of 4 * · 17% of its score · 1 test
Trails Voxtral Mini 4B Realtime (2602) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Voxtral report · English long-form 4-set WER lab | 7.1% ✎ | ▄··█·······▄ beats 3/3 | 17% |
multilingual transcription▂··▃▄█▆▅▄▂·▂ #27 of 36 * · 33% of its score · 2 tests
Trails Omnilingual ASR CTC 1B v2, Omnilingual ASR CTC 7B v2, SeamlessM4T v2 Large here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Voxtral report · Common Voice 12-language error rate lab | 8.1% ✎ | █··█·······█ beats 2/2 | 17% |
| Voxtral report · FLEURS 13-language error rate lab | 5.9% ✎ | █·█········█ beats 2/2 | 16% |
other tests3 not in a task · 47% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Dutch ASR Leaderboard · average WER (FLEURS, VoxPopuli, MLS nl) | 8.2% ✎ | ▄··▄█···█··▄ beats 5/6 | 20% |
| Voxtral report · English short-form 10-set WER lab | 7.3% ✎ | █··█·█·····█ beats 3/3 | 18% |
| koedesk FLEURS first-200 · 9-language headline error mean | 5.2% ✎ | ▄·····▄··▄█▄ beats 1/5 | 9.3% |
$ where to run it
Through an API
api · Mistral AI · $0.0030/audio min · docs.mistral.ai ↗ · price sourceUSD per input audio minute. Exact endpoint mistral/voxtral-mini-2602. Product: Audio transcription. API model: voxtral-mini-2602. Settings: endpoint=/v1/audio/transcriptions, streaming=no. Price basis: published rate. Currency: USD.
$ how the score was worked out · Voxtral Realtime 68%, Dutch ASR Leaderboard (tvosch… 20%, koedesk STT Bench v1 (Track A… 9%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Voxtral Realtime · lab-reported | 4 | 68% |
 Dutch ASR Leaderboard (tvosch) · committed result files | 1 | 20% |
| koedesk STT Bench v1 (Track A, FLEURS first-200) | 1 | 9% |
| prior · pull toward the middle | 3% | |
| weighted average | = 59.3 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json