HumMusQA: leaderboard
Metric: Accuracy (%) on HumMusQA's 320 expert-written four-option music questions (melody, harmony, rhythm, instrumentation, structure, mood, genre and other dimensions), averaged over four runs with the answer options shuffled each run; audio plus question prompt, free-text answers mapped to options by an LLM; chance is 25%; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen2.5-Omni-7B | 64.3 | #601 |
| 2 | Gemini 2.5 Flash | 55.9 | #237 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=hummusqa · How It Works · Data refreshed daily, snapshot 2026-10-11.