Math-PT (pt-BR MC) - Level 5: leaderboard
Metric: Accuracy (%) on the 110 level-5 multiple-choice questions in Brazilian Portuguese (ELLM and ITA); 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: Around January 2027. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 14B | 90.91 |
| 2 | GPT-5 | 90 |
| 3 | Gemini 2.5 Flash | 89.09 |
| 4 | Qwen 3 235B A22B | 89.09 |
| 5 | Qwen 3 32B | 87.27 |
| 6 | DeepSeek V3.1 | 86.36 |
| 7 | Claude Haiku 4.5 | 81.82 |
| 8 | Qwen 3 30B A3B | 80 |
| 9 | Qwen 3 8B | 76.36 |
| 10 | Gemma 3 27B (IT) | 61.82 |
| 11 | Llama 3.3 70B Instruct | 31.82 |
Interactive version: theaggregate.ai/benchmark?slug=math-pt-pt-br-mc-level-5 · How It Works · Data refreshed daily, snapshot 2026-10-07.