MuseBench - Game Arts (Single-Select): leaderboard
MuseBench evaluates intent-level audiovisual-arts understanding in multimodal models over 4,016 questions from video essays in four art categories (cinematic, static visual, stage performing, game arts), with single-select and multi-select framings. This board reports Game-arts chance-adjusted accuracy (%): raw single-select accuracy minus the per-item chance baseline, rescaled so random guessing scores 0 and a correct answer 1; a negative value is worse than chance; higher is better.
Metric: Game-arts chance-adjusted accuracy (%): raw single-select accuracy minus the per-item chance baseline, rescaled so random guessing scores 0 and a correct answer 1; a negative value is worse than chance; higher is better. Source: arxiv.org. Saturation forecast: Around February 2028. 28 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 Plus | 38.8 | #232 |
| 2 | Gemini 3.1 Pro (Preview) | 38.72 | #94 |
| 3 | Qwen 3.5 397B A17B | 35.79 | #246 |
| 4 | Claude Opus 4.6 | 34.07 | #102 |
| 5 | GPT-5.4 | 32 | #123 |
| 6 | InternVL3-78B | 31.59 | #738 |
| 7 | Gemma 4 E4B | 10.3 | #2214 |
| 8 | Qwen2.5-Omni-7B | 8.31 | #1789 |
| 9 | InternVL3-8B | 6.66 | #1688 |
| 10 | Grok 4.1 | 3.2 | #476 |
| 11 | Kimi K2.5 | 0.35 | #256 |
| 12 | GLM-4.5V | -4.34 | #744 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=musebench-game-arts-single-select · How It Works · Data refreshed daily, snapshot 2026-11-13.