OpenAI: whisper-tiny
2 benchmarks
Leaderboard position
#673
Overall rank (of 1121 models)
723
Elo rating
2
Benchmarks
Top 60.0%
Percentile
In OpenAI (of 74 models)
Above: Yandex: gpt-5-pro #672
Below: nace-ai: drex-v1.5 #674
Benchmark highlights
Best benchmark
LibriSpeech (test-clean) (-)
-7.6
Worst benchmark
LibriSpeech (test-other) (-)
-16.9
Nearest competitor from another company
The model from another company with the most similar benchmark scores.
appeared as nearest in 1 of 2 benchmarks
1/2
50.0%
1/2
50.0%
| Model | Votes | % of benchmarks |
|---|---|---|
| cactus-compute: whistle | 1 | 50.0% |
| moonshine-ai: moonshine-tiny | 1 | 50.0% |
Benchmarks
Predicted scores
Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.
| Benchmark | Predicted | Confidence | Based on |
|---|---|---|---|
| MATH-500 | ~84.1 | 50% | — |
| AIME 2025 | ~61.79 | 50% | — |
| AIME 2026 | ~70.95 | 50% | — |
| HMMT Feb 2026 | ~62.45 | 50% | — |
| GPQA Diamond | ~55.19 | 50% | — |
| MMLU-Pro | ~58 | 50% | — |
| Humanity's Last Exam (no tools) | ~9.3 | 50% | — |
| IFEval | ~80.48 | 50% | — |
| IFBench | ~51.16 | 50% | — |
| HumanEval | ~69.89 | 50% | — |
| MBPP | ~72.04 | 50% | — |
| Terminal-Bench v2.1 | ~16.95 | 50% | — |
| MMLU | ~64.29 | 50% | — |
| SWE-bench Verified | ~35.07 | 50% | — |
| SWE-bench Pro | ~14.29 | 50% | — |
| SciCode | ~24.87 | 50% | — |
| AA-LCR | ~40.53 | 50% | — |
| Artificial Analysis Intelligence Index v5.0.1 | ~12 | 50% | — |
| Humanity's Last Exam (with tools) | ~12.73 | 50% | — |
| GSM8K | ~77.88 | 50% | — |
| -(CoderPPL) | ~-1.74 | 50% | — |
| LiveCodeBench v6 | ~52.93 | 50% | — |
| MMMU-Pro | ~50.49 | 50% | — |
| MATH | ~55.24 | 50% | — |
| BBH | ~62.41 | 50% | — |