whatmodel.sh / stt / Chirp 2
Chirp 2
↗ share
+ contribute
Chirp 2 is a closed-weights speech-to-text model from Google. Licence: proprietary.
Catalogue identity notes
Google Cloud Speech-to-Text v2 model 'chirp_2' (leaderboard entry google/chirp_2). Distinct from catalogue google-chirp-3 and from the original 'chirp'. Google updated the GA chirp_2 model on 2024-10-07; the leaderboard does not state which date it was run, so no release month is given.
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 Chirp 2, out of 176 with a score, then more from Google.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 53 |  Universal-Streaming (English) | 58±3 | api $0.0025/audio min | ||
| 54 |  Parakeet RNNT 1.1B | 58±2 | local | ||
| 55 | Chirp 2 | 58±3 | api price unknown | ||
| 56 | Canary-1B-v2 | 58±1 | local | ||
| 57 |  Phi-4-multimodal-instruct | 58±2 | local | ||
| more from Google | |||||
| 11 |  Gemini 3.5 Transcribe | 67±6 | api $0.0050/audio min | ||
| 31 | Chirp 3 | 61±2 | api price unknown | ||
| 33 | Gemini 3.5 Transcribe Live | 61±3 | api $0.0090/audio min | ||
see all speech-to-text models · find the ones that fit your machine
$ what it's good at
other tests5 not in a task · 97% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Bangla STT benchmark · BanSpeech 1,001-clip batch CER | 21.9% ✎ | █·▄··▄··▄·▄▄ beats 2/6 | 37% |
| Open ASR English short-form WER (March 2026) | 6.4% ✎ | ▁▁▁▃··▁▁▂██▆ beats 71/86 | 29% |
| verli STT benchmark · LibriSpeech test-clean WER | 2.3% ✎ | ▃·····▃▅█▃▅▅ beats 2/11 | 17% |
| verli STT benchmark · SPGISpeech 1,200-clip sample WER | 2.3% ✎ | ▂·····▄▂█▂▃▄ beats 9/11 | 7.9% |
| verli STT benchmark · Earnings-22 WER (2,741 utt.) | 12.7% ✎ | ▃·▃▅··▃▃▅▅█▃ beats 4/13 | 7.2% |
$ where to run it
Through an API
api · Google Cloud · price unknown · cloud.google.com ↗
$ how the score was worked out · Bangla Speech-to-Text Benchma… 37%, verliapp/stt-benchmark · inde… 32%, Open ASR Leaderboard 29%
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 |
|---|---|---|
 Bangla Speech-to-Text Benchmark v1.0 (Riverborn) | 1 | 37% |
| verliapp/stt-benchmark · independent STT accuracy benchmark | 3 | 32% |
| Open ASR Leaderboard | 1 | 29% |
| prior · pull toward the middle | 3% | |
| weighted average | = 58.0 |
$ 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