MathArena - Project Euler 943-970 — leaderboard

Metric: Accuracy (%, direct Project Euler problems 943-970). Source: matharena.ai. 18 models tracked.

Top models

#ModelScore
1Human Expert100
2GPT-5.4 (xHigh)87.5
3Gemini 3.1 Pro (Preview)85.71
4Claude Opus 4.6 (High)83.93
5GPT-5.2 (High)80.36
6Gemini 3.5 Flash78.57
7GLM-5.167.86
8GPT-5.1 (High)66.07
9Gemini 3 Pro (Preview)64.29
10Kimi K2.5 (Thinking)63.39
11Gemini 3 Flash60.71
12GPT-5 (High)58.04
13O4 Mini (High)48.21
14DeepSeek V3.2 (Thinking)47.32
15Grok 446.43

Interactive version: theaggregate.ai/benchmark?slug=matharena-project-euler-943-970 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.