LLM Stats (MMAU): leaderboard
MMAU evaluates audio understanding and reasoning across speech, environmental sounds and music using natural-language questions about audio clips. LLM Stats mirrors reported accuracy as percentages; the parent board does not identify test-mini versus full-test runs.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Around February 2027. 5 models tracked. Updated 6 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen3.8 Omni Flash | 81.8 | #127 |
| 2 | Inkling | 77.2 | #279 |
| 3 | Inkling Small | 77 | #309 |
| 4 | Nova 2 Omni | 75.3 | #431 |
| 5 | Qwen2.5-Omni-7B | 65.6 | #1884 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-mmau · How It Works · Data refreshed daily, snapshot 2026-11-21.