Skip to content

Benchmark

MMAU Sound

Scores are percentages.

Models measured
6
Scores recorded
11
Sources
2
IQ weight
0.71

Score distribution

68.2%
Mean
68.9%
Median
4.1%
Std dev
62.7%
10th percentile
73.3%
90th percentile
62.5%
Min
62.5%75.1%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.71 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Gemini 2.5 ProTest-mini· MMAU (audio) data75.1%
2Gemini 2.5 FlashTest-mini· MMAU (audio) data73.3%
3Gemini 2.0 FlashTest-mini· MMAU (audio) data71.2%
4Gemini 2.5 ProTest· MMAU (audio) data70.6%
5Gemini 2.5 FlashTest· MMAU (audio) data69.5%
6Gemini 2.0 FlashTest· MMAU (audio) data68.9%
7Qwen2.5-Omni 7B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim67.9%
8Phi 4 Multimodal InstructTest-mini· MMAU (audio) data65.5%
9Gemini 2.5 Flash-LiteTest-mini· MMAU (audio) data63.1%
10Phi 4 Multimodal InstructTest· MMAU (audio) data62.7%
11Gemini 2.5 Flash-LiteTest· MMAU (audio) data62.5%

Data credits: Data by LLM Stats