whatmodel.sh / stt / Distil-Whisper distil-large-v2
Distil-Whisper distil-large-v2
↗ share
+ contribute
Distil-Whisper distil-large-v2 is an open-weights speech-to-text model from Hugging Face (Distil-Whisper), released October 2023. Its published size is 0.76B parameters. Licence: mit.
Catalogue identity notes
Exact checkpoint distil-whisper/distil-large-v2; distinct from catalogue distil-large-v3-5. Params: HF safetensors total. released: HF repo creation month.
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 Distil-Whisper distil-large-v2, out of 176 with a score, then more from Hugging Face (Distil-Whisper).
| # | model | score | where | ||
|---|---|---|---|---|---|
| 120 |  Ultravox v0.6 (Llama 3.1 8B) | 49±5 | comparison only | ||
| 121 |  Whisper medium.en | 48±3 | local | ||
| 122 | Distil-Whisper distil-large-v2 | 48±2 | local | ||
| 123 | Distil-Whisper distil-small.en | 48±4 | local | ||
| 124 |  CrisperWhisper | 47±3 | local | ||
| more from Hugging Face (Distil-Whisper) | |||||
| 60 | Distil-Whisper distil-large-v3.5 | 58±1 | local | ||
| 77 | Distil-Whisper distil-large-v3 | 56±2 | local | ||
| 138 | Distil-Whisper distil-medium.en | 44±3 | local | ||
see all speech-to-text models · find the ones that fit your machine
$ what it's good at
long-form transcription▁▁▄▁·▁▂█▄▁▁▁ #22 of 27 * · 16% of its score · 1 test
Trails STT En Conformer-Transducer Small here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Open ASR English long-form WER | 14.7% ✎ | ▁▁▅▁·▃▅█▆▁▂▂ beats 5/30 | 16% |
other tests7 not in a task · 82% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Open ASR English short-form WER (March 2026) | 7.9% ✎ | ▁▁▁▃··▁▁▂██▆ beats 38/86 | 16% |
| Meanwhile long-form WER lab | 16.8% ✎ | █·······██·█ beats 0/3 | 14% |
| TED-LIUM long-form WER lab | 9.7% ✎ | █······█·█·█ beats 0/3 | 13% |
| Rev16 long-form WER lab | 21.1% ✎ | █········███ beats 0/3 | 11% |
| Earnings-21 long-form WER lab | 15.1% ✎ | █········███ beats 0/3 | 10% |
| Earnings-22 long-form WER lab | 19.1% ✎ | █······█·█·█ beats 0/3 | 10% |
| Distil-Whisper · 4-set out-of-distribution short-form mean WER lab | 10.1% ✎ | █·█··█·█···█ beats 2/4 | 7.5% |
$ where to run it
Other ways to run it
- Run it as published · needs ~2.32 GB · from huggingface.co/distil-whisper (official)
uv run --with transformers --with torch python -c "from transformers import pipeline; print(pipeline('automatic-speech-recognition', model='distil-whisper/distil-large-v2')('audio.wav', return_timestamps=True)['text'])"
Official weights and documentation: huggingface.co/distil-whisper ↗
$ how the score was worked out · distil-whisper/distil-large-v… 58%, Open ASR Leaderboard 33%, huggingface/distil-whisper RE… 8%
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 |
|---|---|---|
 distil-whisper/distil-large-v3.5 model card · lab-reported | 5 | 58% |
| Open ASR Leaderboard | 2 | 33% |
 huggingface/distil-whisper README (Distil-Whisper headline WER table) · lab-reported | 1 | 8% |
| prior · pull toward the middle | 2% | |
| weighted average | = 47.9 |
$ 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