Math-PT (pt-PT MC) - Level 3: leaderboard
Metric: Accuracy (%) on the 143 level-3 multiple-choice questions in European Portuguese (OPM); native-language zero-shot prompt asking for a boxed final answer, figures given as LaTeX source; the letter in the final boxed answer must match the key; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 96.5 |
| 2 | DeepSeek V3.1 | 88.81 |
| 3 | Qwen 3 235B A22B | 88.11 |
| 4 | Qwen 3 14B | 86.71 |
| 5 | Qwen 3 30B A3B | 86.71 |
| 6 | Qwen 3 32B | 85.31 |
| 7 | Gemini 2.5 Flash | 83.92 |
| 8 | Qwen 3 8B | 78.32 |
| 9 | Claude Haiku 4.5 | 76.92 |
| 10 | Gemma 3 27B (IT) | 48.95 |
| 11 | Llama 3.3 70B Instruct | 23.78 |
Interactive version: theaggregate.ai/benchmark?slug=math-pt-pt-pt-mc-level-3 · How It Works · Data refreshed daily, snapshot 2026-10-07.