KPM-QA - Human: leaderboard
Metric: Accuracy (%) by option matching on KPM-QA human (how many people appear and move, interactions and whether movements differ) questions about motion-centric human videos (answers generated by GPT-4.1 from the video and kinematics-grounded captions built from 3D pose estimation); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4.1 | 77.62 | #240 |
| 2 | Gemini 2.5 Pro | 76.22 | #145 |
| 3 | Qwen 2.5 VL 7B Instruct | 56.56 | #643 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=kpm-qa-human · How It Works · Data refreshed daily, snapshot 2026-10-11.