whatmodel.sh / stt / Granite Speech 5.0 470M TurboCTC NC
Granite Speech 5.0 470M TurboCTC NC
↗ share
+ contribute
Granite Speech 5.0 470M TurboCTC NC is an open-weights speech-to-text model from IBM, released August 2026. Its published size is 0.473B parameters. Licence: cc-by-nc-sa-4.0.
Catalogue identity notes
ibm-granite/granite-speech-5.0-470m-turboctc-nc: research/non-commercial English CTC model (card disclaimer: 'intended for research and non-commercial use-only. See Granite Speech 5.0 Turbo CTC if interested in commercial use'; 'compact 470 million parameter English ASR model'). Different weights from the Apache-2.0 granite-speech-5-0-470m-turboctc (catalogued; transcribe.cpp scores them separately, 1.29% vs 1.34% LibriSpeech). Port pins upstream 0eb7b4f (2026-08-26); later commit af0a8e8 only adds results to the card. Repo created 2026-08-17.
Model information sources
Release date ↗ · Parameter count ↗
Hugging Face safetensors metadata total 472,993,792 (current main af0a8e8); transcribe.cpp catalog counts 473,014,752 for the converted port.
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 5.0 470M TurboCTC NC, out of 176 with a score, then more from IBM.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 13 |  GPT Transcribe | 64±4 | api $0.0045/audio min | ||
| 14 |  Zoom Scribe v1 | 64±5 | api price unknown | ||
| 15 | Granite Speech 5.0 470M TurboCTC NC | 64±4 | local | ||
| 16 |  Cohere Transcribe Arabic (07-2026) | 63±3 | local | ||
| 17 |  Universal-3 Pro | 63±2 | comparison only | ||
| more from IBM | |||||
| 18 | Granite Speech 4.1 2B | 62±3 | local | ||
| 19 | Granite 4.0 1B Speech | 62±3 | local | ||
| 20 | Granite Speech 5.0 470M TurboCTC | 62±4 | local | ||
see all speech-to-text models · find the ones that fit your machine
$ what it's good at
other tests2 not in a task · 94% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LibriSpeech test-clean WER | 1.3% ✎ | ▁···▁▁▂·▄▄██ beats 54/56 | 67% |
| FLEURS English test WER | 4.3% ✎ | ▁··▁▁▁▁▁▁▃█▁ beats 43/56 | 27% |
$ where to run it
Other ways to run it
- Run it as published · needs ~0.51 GB · from github.com/handy-computer
git clone https://github.com/handy-computer/transcribe.cpp.git && cd transcribe.cpp && cmake -B build && cmake --build build && hf download handy-computer/granite-speech-5.0-470m-turboctc-nc-gguf granite-speech-5.0-470m-turboctc-nc-Q8_0.gguf --local-dir models/granite-speech-5.0-470m-turboctc-nc && build/bin/transcribe-cli -m models/granite-speech-5.0-470m-turboctc-nc/granite-speech-5.0-470m-turboctc-nc-Q8_0.gguf audio.wav
Weights and documentation: huggingface.co/ibm-granite ↗
$ how the score was worked out · transcribe.cpp model catalog… 94%
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 |
|---|---|---|
 transcribe.cpp model catalog · LibriSpeech test-clean WER of validated full-precision ports | 2 | 94% |
| prior · pull toward the middle | 6% | |
| weighted average | = 63.5 |
$ 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