Skip to content

LLM Benchmark Hub

Model evidence, daily

Models What's New Trends Benchmarks Skill Maps
How It Works Guesswork Metrics
LLM Benchmark Hub
Loading data...

ErdosBench — leaderboard

Metric: Judge Average (0-4). Source: www.ulam.ai. 7 models tracked.

Top models

#ModelScore
1GPT-5.6 Sol (xHigh)3.12
2GPT-5.5 (xHigh)2.68
3Kimi K2.7 Code2.62
4Claude Opus 4.8 (Max)2.51
5GLM-5.22.4
6Qwen 3.7 Max2.37
7MiniMax-M32.11

Interactive version: aibenchmarks.dev/benchmark?slug=erdosbench · How the rankings work · Data refreshed daily, snapshot 2026-07-20.

Built by Mikhail Doroshenko — AI researcher, co-author of Humanity’s Last Exam. Independent project; no affiliation with any AI lab.