MathNet: leaderboard

MathNet (2026) is a multilingual set of Olympiad problems with official solutions from national competition booklets of 47 countries; on its problem-solving test set a GPT-5 grader scores each solution from 0 to 7 against the official one, and 6 or more counts as correct. Accuracy over all test problems, in percent; higher is better.

Metric: Problem-solving accuracy (%) over all problems (micro average) on MathNet-Solve-Test, Olympiad problems from official national competition booklets of 47 countries; a GPT-5 grader scores each solution 0-7 against the official solution and 6 or more counts as correct; image-capable models get the figures, text-only models a text description of them; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 14 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3.1 Pro (Preview)78.4#99
2Gemini 3 Flash (Preview)70.4#152
3GPT-569.3#178
4GPT-5 Mini57#364
5Claude Opus 4.6Thinking45.7#113
6GPT-5 Nano42.2#906
7Gemini 2.5 Flash41.1#463
8DeepSeek V3.2Non-reasoning40.1#419
9MathNet DeepSeek-R1 (checkpoint unspecified)36.3
10Grok 328.5#611
11GPT-4.121.4#505
12Llama 4 Maverick Instruct FP814.7#1023
13GPT-4o6.8#751
14Ministral 3B4.4#6858

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=mathnet · How It Works · Data refreshed daily, snapshot 2026-11-21.