Benchmark
MMAU
Scores are percentages. Higher is better.
All speech benchmarksBenchmark sourceScores from MMAU (audio) data, BenchBoard docs, LLM Stats data.
- Models and systems measured
- 14
- Scores recorded
- 21
- Sources
- 3
- IQ weight
- 0.00
Score distribution
- 72.4%
- Mean
- 71.6%
- Median
- 6.9%
- Std dev
- 63.5%
- 10th percentile
- 81.8%
- 90th percentile
- 61.6%
- Min
This benchmark has a weight of 0.00 for chat models, measured over 12 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model or system | Score |
|---|---|---|
| 1 | Qwen3.5 Omni PlusQwen3.5-Omni release · published setting· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 82.2% |
| 2 | Qwen3.5 Omni PlusQwen3.8 Omni Flash 官方发布表· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 81.9% |
| 3 | Qwen3.8 Omni Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 81.8% |
| 4 | Qwen3.8 Omni FlashQwen3.8 Omni Flash 官方发布表· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 81.8% |
| 5 | Qwen3.5 Omni FlashQwen3.5-Omni release · published setting· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 80.4% |
| 6 | Inkling FP4· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 77.2% |
| 7 | Inkling Small· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 77% |
| 8 | Gemini 3.8 FlashQwen3.8 Omni Flash 官方发布表· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76.9% |
| 9 | Gemini 2.5 FlashTest-mini· MMAU (audio) data | 71.8% |
| 10 | Gemini 2.5 ProTest-mini· MMAU (audio) data | 71.6% |
| 11 | Gemini 2.0 FlashTest-mini· MMAU (audio) data | 70.5% |
| 12 | Gemini 2.5 ProTest· MMAU (audio) data | 69.4% |
| 13 | Gemini 2.5 FlashTest· MMAU (audio) data | 67.4% |
| 14 | Gemini 2.0 FlashTest· MMAU (audio) data | 67% |
| 15 | Gemini 2.5 Flash-LiteTest-mini· MMAU (audio) data | 66.2% |
| 16 | Phi 4 Multimodal InstructTest-mini· MMAU (audio) data | 65.7% |
| 17 | Qwen2.5-Omni 7B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 65.6% |
| 18 | Muse Spark 1.2Qwen3.8 Omni Flash 官方发布表· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 63.5% |
| 19 | Phi 4 Multimodal InstructTest· MMAU (audio) data | 62.8% |
| 20 | Gemini 2.5 Flash-LiteTest· MMAU (audio) data | 61.6% |
Not ranked
The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.
| Model or system | Score |
|---|---|
| Seed-2.0-LiteQwen3.8 Omni Flash 官方发布表· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 77.2% |
Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Data by LLM Stats