CausalPhys - Perception: leaderboard
Metric: Answer accuracy (0-1 scaled to %) on the Perception domain (association rung: recognizing objects and physical attributes) of the image- and video-based questions with verified answers over real-world physical scenes; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-78B | 62.57 |
| 2 | GPT-4o Mini | 59.83 |
| 3 | Gemini 2.5 Flash | 59.2 |
| 4 | GPT-4o | 58.89 |
| 5 | Claude Sonnet 4 | 58.68 |
| 6 | Qwen 3 VL 32B | 56.89 |
| 7 | Phi-4 Multimodal Instruct | 55.73 |
| 8 | Qwen 2 VL 7B | 52.05 |
| 9 | Mistral Small 3.2 | 48.26 |
Interactive version: theaggregate.ai/benchmark?slug=causalphys-perception · How It Works · Data refreshed daily, snapshot 2026-09-29.