Prim - Generation: leaderboard
Prim's direct-solving score: 182 expert-verified free-form mathematics problems from Humanity's Last Exam, answered zero-shot with chain-of-thought and scored by exact match on the final answer; higher is better (Prim, 2026).
Metric: Generation accuracy (%): exact match of the model's own zero-shot chain-of-thought final answer on 182 HLE-Verified mathematics problems; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 12 models tracked. Updated 4 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-5.4 | High | 67.03 | #138 |
| 2 | Qwen 3.6 27B | 52.75 | #352 | |
| 3 | GPT-5.4 Mini | High | 50.55 | #405 |
| 4 | Qwen 3.5 27B | 47.8 | #378 | |
| 5 | GPT-5.4 Nano | High | 44.51 | #651 |
| 6 | GPT-OSS-20B | High | 43.41 | #1166 |
| 7 | Qwen 3.5 9B | 38.46 | #829 | |
| 8 | Qwen 3.5 4B | 26.37 | #1235 | |
| 9 | DeepSeek R1 Distill Qwen 32B | 18.13 | #1813 | |
| 10 | DeepSeek R1 0528 Qwen3 8B | 17.03 | #2190 | |
| 11 | DeepSeek R1 Distill Qwen 14B | 15.93 | #2706 | |
| 12 | DeepSeek-R1-Distill-Qwen-7B | 13.19 | #7912 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=prim-generation · How It Works · Data refreshed daily, snapshot 2026-11-20.