PolyReal - Foundational Knowledge Application: leaderboard
Metric: F1 score on the Foundational Knowledge Application module of PolyReal (applying core scientific principles to real scenarios); higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 | 0.64 |
| 2 | GPT-5 | 0.59 |
| 3 | Claude Sonnet 4.5 (Thinking) | 0.58 |
| 4 | Gemini 2.5 Pro | 0.57 |
| 5 | Grok 4 | 0.57 |
| 6 | Qwen 3 VL 235B A22B (Thinking) | 0.57 |
| 7 | Qwen 3 VL 32B (Thinking) | 0.55 |
| 8 | DeepSeek R1 | 0.55 |
| 9 | Gemini 2.5 Flash | 0.53 |
| 10 | Gemini 2.0 Flash (Thinking) | 0.51 |
| 11 | GPT-4o | 0.39 |
| 12 | Qwen 2.5 VL 72B Instruct | 0.37 |
| 13 | GPT-4o Mini | 0.35 |
Interactive version: theaggregate.ai/benchmark?slug=polyreal-foundational-knowledge-application · How It Works · Data refreshed daily, snapshot 2026-10-07.