handy models

speech-to-text benchmarks for transcribe.cpp, the engine behind Handy

best model by language

Benchmarked on the FLEURS test split at Q8_0. Error is WER — lower is better. Speed is × realtime on Handy target hardware.

accuracy vs speed

Finnish · 17 models

AMD Ryzen 4750U GPU (Vulkan)
pareto frontier (no model is both faster and more accurate) other models
highlighted rows are recommended picks or on pareto frontier
1 whisper-large-v3 7.73 2.0× 1.0× 1.55 GB Apache-2.0
2 whisper-large-v3-turboaccurate pick 8.29 3.1× 1.4× 0.83 GB Apache-2.0
3 canary-1b-v2!overall pickfast pick 8.86 14.1× 8.3× 1.07 GB CC-BY-4.0
4 whisper-large-v2 9.58 2.0× 1.0× 1.55 GB Apache-2.0
5 parakeet-ultra 11.14 23.0× 13.3× 0.69 GB CC-BY-4.0
6 orukeet 11.88 25.1× 13.6× 0.69 GB CC-BY-SA-4.0
7 whisper-large 12.73 2.0× 1.0× 1.55 GB Apache-2.0
8 parakeet-tdt-0.6b-v3 13.30 25.1× 13.7× 0.69 GB CC-BY-4.0
9 parakeet-primeline 13.39 25.0× 13.6× 0.69 GB CC-BY-4.0
10 whisper-medium 14.67 4.1× 1.9× 0.77 GB Apache-2.0
11 nemotron-3.5-asr-streaming-0.6b 21.91 17.1× 10.6× 0.70 GB OpenMDW-1.1
12 Qwen3-ASR-1.7B 25.48 4.1× 2.5× 2.04 GB Apache-2.0
13 whisper-small 26.48 11.7× 5.9× 0.25 GB Apache-2.0
14 Qwen3-ASR-0.6B 46.49 9.7× 5.3× 0.79 GB Apache-2.0
15 whisper-base 49.46 29.0× 19.3× 0.08 GB Apache-2.0
16 Fun-ASR-MLT-Nano-2512! 68.16 8.8× 5.5× 0.83 GB FunASR Model Open Source License Agreement v1.1
17 whisper-tiny 68.58 43.1× 39.8× 0.04 GB Apache-2.0