STEMVerse - Chemistry - Physical Chemistry: leaderboard
Metric: Few-shot accuracy (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen3-14B-Base | 40.2 |
| 2 | Qwen3-4B-Base | 37.25 |
| 3 | Qwen 2.5 7B Instruct | 33.33 |
| 4 | Qwen 2.5 7B | 33.33 |
| 5 | Qwen3-8B-Base | 32.35 |
| 6 | Qwen 2.5 14B Instruct | 29.41 |
| 7 | Qwen 2.5 14B | 28.43 |
| 8 | Qwen 2.5 3B | 25.49 |
| 9 | Llama 3.1 8B Instruct | 23.53 |
| 10 | Llama 3.2 3B Instruct | 21.57 |
| 11 | Llama 3.1 8B | 21.57 |
| 12 | Qwen 2.5 3B Instruct | 20.59 |
| 13 | Llama 3.2 3B | 11.76 |
Interactive version: theaggregate.ai/benchmark?slug=stemverse-chemistry-physical-chemistry · How It Works · Data refreshed daily, snapshot 2026-09-25.