Category ranking · 8 models · checked 2026-10-13
Speech to text
Ranked by Open ASR Leaderboard, average WER (English short-form) (lower is better). Source: hf-audio/open-asr-leaderboard-results, english_short_latest.csv (leaderboard version 09-10-2026).
Open ASR Leaderboard, average WER (English short-form)
7 with a published score, 1 not yet measured.
# | Model | Parameters | Score | Source | Licence |
|---|---|---|---|---|---|
| 1 | Canary-Qwen 2.5B NVIDIA | 2.6 B | 6.37%Sourced | Source avg 6.3686, RTFx 881.67. | CC-BY-4.0 |
| 2 | Parakeet TDT 0.6B v2 NVIDIA | — | 6.43%Sourced | Source avg 6.4329, RTFx 6148.32. | CC-BY-4.0 |
| 3 | Parakeet TDT 0.6B v3 NVIDIA | 627 M | 6.51%Sourced | Source avg 6.5143, RTFx 6227.73. | CC-BY-4.0 |
| 4 | Distil-Whisper large-v3.5 Hugging Face | 756 M | 7.02%Sourced | Source avg 7.0229, RTFx 898.21. | MIT |
| 5 | Voxtral Mini 3B Mistral AI | 4.7 B | 7.84%Sourced | Source avg 7.8371, RTFx 179.78. | Apache-2.0 |
| 6 | Whisper large-v3 OpenAI | 1.5 B | 7.84%Sourced | Source avg 7.8414, RTFx 470.73. | Apache-2.0 |
| 7 | Whisper large-v3 turbo OpenAI | 809 M | 8.39%Sourced | Source avg 8.3871, RTFx 810.5. | MIT |
| — | Canary 1B Flash NVIDIA | 811 M | Not yet measured | Not listed in the current leaderboard file. | CC-BY-4.0 |
How to read this: one public benchmark per category, so you can compare like with like. We copy the published number and link it; we do not re-run it, average different benchmarks or fill gaps. Vendor-reported numbers use each vendor's own settings, so treat small gaps as ties. “Not yet measured” means we found no comparable public score, not that the model is weak.
Every model was checked against the Hugging Face API on 2026-10-13: it exists, is public, and has the parameter count and licence shown. Other categories: Math and reasoning · Coding · Vision-language · OCR and documents · Text to speech · Image generation · Embeddings and rerankers · Small fast 'System 1' · Biomedical · Tool use and agents · Multilingual and translation · Time series · Robotics (VLA).
Machine-readable: /api/v1/categories/speech-to-text.json