PIQA — leaderboard
Physical Intuition QA: 20K multiple-choice questions testing physical commonsense reasoning - understanding how everyday objects and actions work in the physical world.
Metric: Accuracy (%). Source: epoch.ai. Status: saturation imminent. 60 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Median Human | 95 |
| 2 | GPT-4o Mini (2024-07-18) | 88.7 |
| 3 | Phi-3.5-MoE-instruct | 88.6 |
| 4 | Gemini 1.5 Flash (002) | 87.5 |
| 5 | Llama 3.1 405B | 85.9 |
| 6 | falcon-180B | 84.9 |
| 7 | DeepSeek V2 | 83.9 |
| 8 | Gemma 2 9B | 83.7 |
| 9 | Mixtral 8x7B (v0.1) | 83.6 |
| 10 | Mistral-Nemo-Base-2407 | 83.5 |
| 11 | StableBeluga2 | 83.3 |
| 12 | Mistral-7B-v0.1 | 83 |
| 13 | falcon-40B | 83 |
| 14 | Llama 2 70B Base | 82.8 |
| 15 | LLaMA-65B | 82.8 |
Interactive version: theaggregate.ai/benchmark?slug=piqa · How the rankings work · Data refreshed daily, snapshot 2026-07-22.