MuseBench: Benchmarking Intent-Level Audiovisual Arts Understanding in MLLMs: leaderboard
Metric: Overall ACC (%). Source: benchmarklist.com. 28 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 | 48.29 |
| 2 | Qwen 3.5 Plus | 47.27 |
| 3 | Qwen 3.5 397B A17B | 44.76 |
| 4 | GPT-5.4 | 44.58 |
| 5 | InternVL3-78B | 37.81 |
| 6 | Gemini 3.1 Pro (Preview) | 36.89 |
| 7 | InternVL3-8B | 33.07 |
| 8 | Gemma 4 E4B | 27.61 |
| 9 | Grok 4.1 | 20.54 |
| 10 | Kimi K2.5 | 19.91 |
| 11 | GLM-4.5V | 17.13 |
Interactive version: theaggregate.ai/benchmark?slug=musebench-benchmarking-intent-level-audiovisual-arts-understanding-in-mllms · How It Works · Data refreshed daily, snapshot 2026-09-19.