VURB - Video Reasoning: leaderboard
Video reasoning split of VURB (Wei et al., 2026): 858 answer pairs to knowledge, spatial and complex-reasoning questions from MMVU, Video-MMLU, VSI-Bench and VideoReasonBench, judged pairwise eight times with the answer order shuffled; a pair counts when the correct answer wins more than four judgments. Accuracy in percent over pairs with a valid verdict; higher is better.
Metric: Majority-vote pairwise accuracy (%; eight judgments with the response order shuffled, a pair counts when the chosen response wins more than four, over pairs with a valid verdict; 858 pairs from MMVU, Video-MMLU, VSI-Bench and VideoReasonBench). Source: arxiv.org. Saturation forecast: Around February 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 63.1 | #172 |
| 2 | Qwen 3 VL 32B (Thinking) | 53.5 | #557 |
| 3 | Qwen 3 VL 8B (Thinking) | 52.8 | #887 |
| 4 | Qwen 3 VL 8B Instruct | 51 | #982 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=vurb-video-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-12.