whatmodel.sh / stt / Orukeet

Orukeet

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

Orukeet is an open-weights speech-to-text model from Oruk AI, released September 2026. Its published size is 0.627B parameters. Licence: cc-by-sa-4.0.

Catalogue identity notes

oruk/orukeet: 25-language recogniser fine-tuned from nvidia/parakeet-tdt-0.6b-v3 (card front matter base_model nvidia/parakeet-tdt-0.6b-v3, base_model_relation finetune; 'replaces half of the encoder's temporal depthwise filters with 12,288 fitted, frozen Gabor kernels'). Card: 'NeMo, ONNX INT8, native Q8 and native F16 all derive from the same r3 release checkpoint (031c8ddab484)'. Measured port pins upstream commit b59c13a (2026-09-24, current main). Repo created 2026-09-09.

Model information sources

Release date ↗ · Parameter count ↗

Hugging Face safetensors metadata total 627,057,286 (current main b59c13a); transcribe.cpp catalog counts 627,052,166 for the converted port.

56.5±3.9#61 of 162 speech-to-text modelsfrom 2 benchmark results across 1 source* thinly tested, treat the rank as provisional

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 Orukeet, out of 176 with a score.

#modelscorewhereRAMreleased
69Parakeet RNNT 0.6B NVIDIA57±3local~3.5GB2024-01
70Rev AI Fusion (Whisper Fusion) Rev57±4comparison only2024-01
71Orukeet Oruk AI57±4local~0.7GB2026-09
72Zipformer Transducer XL 290M SoundsGood AI56±5comparison only2026-05
73Parakeet TDT-CTC 1.1B NVIDIA56±4local~1.3GB2024-05

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
testresultvs the others in its rundifficultycountswhy
LibriSpeech test-clean WER1.9% ✎▁···▁▁▂·▄▄██ beats 33/56avg 2.3%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 67%blunt
FLEURS English test WER4.4% ✎▁··▁▁▁▁▁▁▃█▁ beats 41/56avg 5.8%▇▇▇▇▇▇▇▇▇▇▇ 27%blunt

$ where to run it

Other ways to run it

  1. Run it as published · needs ~0.74 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/orukeet-gguf orukeet-Q8_0.gguf --local-dir models/orukeet && build/bin/transcribe-cli -m models/orukeet/orukeet-Q8_0.gguf audio.wav

Weights and documentation: huggingface.co/oruk ↗

$ 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

sourceresultsshare of the scorelicence
transcribe.cpp model catalog · LibriSpeech test-clean WER of validated full-precision ports2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 94%MIT
prior · pull toward the middle▇▇ 6%
weighted average= 56.5

$ 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