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
Vietnamese · 15 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 | Qwen3-ASR-1.7Baccurate pick | 6.15 | 4.1× | 2.5× | 2.04 GB | Apache-2.0 |
| 2 | cohere-transcribe-03-2026!overall pick | 7.39 | 8.1× | 4.3× | 2.25 GB | Apache-2.0 |
| 3 | Fun-ASR-MLT-Nano-2512! | 8.32 | 8.8× | 5.5× | 0.83 GB | FunASR Model Open Source License Agreement v1.1 |
| 4 | whisper-large-v3 | 8.74 | 2.0× | 1.0× | 1.55 GB | Apache-2.0 |
| 5 | Qwen3-ASR-0.6B | 9.32 | 9.7× | 5.3× | 0.79 GB | Apache-2.0 |
| 6 | whisper-large-v3-turbo | 9.48 | 3.1× | 1.4× | 0.83 GB | Apache-2.0 |
| 7 | moonshine-base-vifast pick | 9.96 | 23.9× | 15.0× | 0.07 GB | MIT |
| 8 | whisper-large-v2 | 11.25 | 2.0× | 1.0× | 1.55 GB | Apache-2.0 |
| 9 | whisper-large | 11.51 | 2.0× | 1.0× | 1.55 GB | Apache-2.0 |
| 10 | moonshine-tiny-vi | 13.37 | 39.5× | 27.0× | 0.03 GB | MIT |
| 11 | whisper-medium | 13.74 | 4.1× | 1.9× | 0.77 GB | Apache-2.0 |
| 12 | nemotron-3.5-asr-streaming-0.6b | 13.96 | 17.1× | 10.6× | 0.70 GB | OpenMDW-1.1 |
| 13 | whisper-small | 22.47 | 11.7× | 5.9× | 0.25 GB | Apache-2.0 |
| 14 | whisper-base | 42.60 | 29.0× | 19.3× | 0.08 GB | Apache-2.0 |
| 15 | whisper-tiny | 64.49 | 43.1× | 39.8× | 0.04 GB | Apache-2.0 |