MuseBench - Cinematic Arts (Multi-Select): leaderboard
Metric: Cinematic-arts exact-match accuracy (%) on multi-select questions; higher is better. Source: arxiv.org. Saturation forecast: Around 2034. 28 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 397B A17B | 33.33 |
| 2 | GPT-5.4 | 31.58 |
| 3 | Claude Opus 4.6 | 31.58 |
| 4 | Qwen 3.5 Plus | 29.04 |
| 5 | InternVL3-8B | 28.07 |
| 6 | Qwen2.5-Omni-7B | 21.44 |
| 7 | Gemini 3.1 Pro (Preview) | 15.98 |
| 8 | GLM-4.5V | 14.62 |
| 9 | InternVL3-78B | 14.23 |
| 10 | Gemma 4 E4B | 11.89 |
| 11 | Grok 4.1 | 6.63 |
| 12 | Kimi K2.5 | 2.73 |
Interactive version: theaggregate.ai/benchmark?slug=musebench-cinematic-arts-multi-select · How It Works · Data refreshed daily, snapshot 2026-09-29.