MOV-Bench - Relational Reasoning: leaderboard
The 127 relational-reasoning questions of MOV-Bench (Xu et al., 2026), asking how people or objects relate from audio-visual evidence spread across several segments of a FineVideo clip. Accuracy in percent when the omni-modal model answers directly from the whole video; higher is better.
Metric: Accuracy (%; 127 relational-reasoning questions; direct inference on the whole video). Source: arxiv.org. Saturation forecast: Around February 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen2.5-Omni-7B | 55.12 | #1797 |
| 2 | Qwen3 Omni 30B A3B Instruct | 54.33 | #902 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=mov-bench-relational-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-12.