Benchmark
MMAU-Pro Speech–Sound
Scores are percentages.
- Models measured
- 11
- Scores recorded
- 11
- Sources
- 1
- IQ weight
- 0.81
Score distribution
- 59.5%
- Mean
- 60.2%
- Median
- 15.2%
- Std dev
- 31.8%
- 10th percentile
- 76.1%
- 90th percentile
- 30.6%
- Min
This benchmark has a weight of 0.81 for chat models, measured over 9 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro Preview· MMAU-Pro data | 76.1% |
| 2 | Gemini 3.7 Flash· MMAU-Pro data | 76.1% |
| 3 | Gemini 3.5 Flash· MMAU-Pro data | 75% |
| 4 | Gemini 3.6 Flash· MMAU-Pro data | 70.5% |
| 5 | Gemini 2.5 Flash· MMAU-Pro data | 61.3% |
| 6 | Qwen2.5-Omni 7B· MMAU-Pro data | 60.2% |
| 7 | qwen3-omni-30b-a3b-instruct· MMAU-Pro data | 59.1% |
| 8 | GPT Audio· MMAU-Pro data | 58% |
| 9 | Gemini 2.0 Flash· MMAU-Pro data | 55.9% |
| 10 | Gemma 3n E4B IT· MMAU-Pro data | 31.8% |
| 11 | Gemma 3n E2b It· MMAU-Pro data | 30.6% |