Whisper Large-v3: benchmark results
Provider: Other. Released 2023-11-06. Access: Open.
Unified ELO 1422 ± 31, rank #1729 of 2656 rated models, from 30 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AudioBench - GigaSpeech WER | 9.46 | WER (%) (lower is better) | 100 |
| AudioBench - IMDA Part 3 ASR WER | 27.03 | WER (%) (lower is better) | 100 |
| AudioBench - LibriSpeech Other WER | 3.66 | WER (%) (lower is better) | 100 |
| AudioBench - TED-LIUM 3 Long Form WER | 3.21 | WER (%) (lower is better) | 100 |
| ProLLM - Transcription | 77.9 | Score (%) | 100 |
| AudioBench - IMDA Part 5 ASR WER | 21.44 | WER (%) (lower is better) | 88.9 |
| AudioBench - TED-LIUM 3 WER | 3.76 | WER (%) (lower is better) | 88.9 |
| AudioBench - YouTube ASR Batch 2 WER | 17.21 | WER (%) (lower is better) | 87.5 |
| AudioBench - SEAME Singapore English WER | 53.77 | WER (%) (lower is better) | 80 |
| AudioBench - Earnings21 WER | 11.86 | WER (%) (lower is better) | 77.8 |
| AudioBench - Earnings22 WER | 15.89 | WER (%) (lower is better) | 77.8 |
| AudioBench - IMDA Part 1 ASR WER | 6.84 | WER (%) (lower is better) | 77.8 |
Interactive version: theaggregate.ai/model?slug=whisper-large-v3 · How It Works · Data refreshed daily, snapshot 2026-09-19.