PECC - Project Euler: leaderboard

The Project Euler track of PECC (arXiv) — math-heavy Project Euler problems solved via generated code; solve rate (%).

Metric: Solve rate (%). Source: huggingface.co. Saturation forecast: Around July 2027. 11 models tracked.

Top models

#ModelScoreOverall rank
1GPT-3.5 Turbo8.19#865
2Llama 3 8B Instruct4.47#1129
3Phi-3 Mini 128K Instruct3.35#1298
4Mixtral 8x7B Instruct (v0.1)2.86#1073
5WizardLM-2-7B1.24#1395
6Mistral Instruct v0.1 7B0.37#1454

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=pecc-project-euler · How It Works · Data refreshed daily, snapshot 2026-10-20.