whatmodel.sh / stt / Granite Speech 4.1 2B

Granite Speech 4.1 2B

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Granite Speech 4.1 2B is an open-weights speech-to-text model from IBM, released April 2026. Its published size is 2.31B parameters. Licence: apache-2.0.

62.3±3.3#15 of 162 speech-to-text modelsfrom 3 benchmark results across 2 sources

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 Granite Speech 4.1 2B, out of 176 with a score, then more from IBM.

#modelscorewhereRAMreleased
16Cohere Transcribe Arabic (07-2026) Cohere63±3local~4.1GB2026-07
17Universal-3 Pro AssemblyAI63±2comparison only2026-02
18Granite Speech 4.1 2B IBM62±3local~6GB2026-04
19Granite 4.0 1B Speech IBM62±3local~6GB2026-03
20Granite Speech 5.0 470M TurboCTC IBM62±4local~1.6GB2026-08
more from IBM
15Granite Speech 5.0 470M TurboCTC NC IBM64±4local~0.5GB2026-08
26Granite Speech 4.1 2B NAR IBM62±4local~2.5GB2026-03
34Granite Speech 4.1 2B Plus IBM61±4local~2.4GB2026-04

see all speech-to-text models · find the ones that fit your machine

$ what it's good at

other tests4 not in a task · 96% of its score
testresultvs the others in its rundifficultycountswhy
LibriSpeech test-clean WER1.3% ✎▁···▁▁▂·▄▄██ beats 53/56avg 2.3%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 48%blunt
AfriCodeSwitch English-Yoruba code-switched speech · WER (cap2000)73.5% ✎█··█·█···█·█ beats 0/4avg 131%▇▇▇▇▇▇▇▇▇▇▇ 28%typical
FLEURS English test WER4.1% ✎▁··▁▁▁▁▁▁▃█▁ beats 49/56avg 5.8%▇▇▇▇▇▇▇▇ 19%blunt
Granite Speech · LScln punctuation error lab25.7% ✎no other model yet—not counted yet: no other model in the same run to compare it with

$ where to run it

Other ways to run it

version

  1. Run it as published · needs ~6.04 GB, Mac only · from huggingface.co/ibm-granite
    uvx --from mlx-audio mlx_audio.stt.generate --model ibm-granite/granite-speech-4.1-2b --audio audio.wav --output-path transcript

Weights and documentation: huggingface.co/ibm-granite ↗

$ how the score was worked out · transcribe.cpp model catalog… 68%, Switch-aware evaluation of AS… 28%

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

sourceresultsshare of the scorelicence
transcribe.cpp model catalog · LibriSpeech test-clean WER of validated full-precision ports2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 68%MIT
Switch-aware evaluation of ASR on English-Yoruba code-switched speech (arXiv:2609.11786v1)1▇▇▇▇▇▇▇▇ 28%CC BY 4.0
prior · pull toward the middle▇ 4%
weighted average= 62.3

not counted (1)

IBM Granite Speech 4.1 own punctuation evaluationGranite Speech · LScln punctuation error [correct]25.7%fewer than three models have this test yet, or this source ran it for this model alone

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

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