Epoch AI runs AI models itself on a set of hard tests (maths, science, coding) and publishes the results in one place, so you can compare models on tests the labs did not grade themselves.
Leaderboard
Top 3 right now
1stclaude-fable-5-2_max95.1
2ndgpt-6.1-sol_max93.7
3rdgpt-6-astra_max93.7
What it measures
Independently run and collected results on many benchmarks (such as FrontierMath, GPQA Diamond, SWE-bench) with trends over time.
How to read the numbers
Share of FrontierMath Tiers 1-3 (v2, private set) maths problems solved (%) — higher is better.
Note: The Hub covers dozens of benchmarks; only one representative table is shown here (FrontierMath Tiers 1-3 v2, Epoch's 'best score across scorers', converted from 0-1 to %). Sorted by score; the file itself is not sorted. Data zip generated 2026-11-17.
Source: Epoch AI · epoch.ai · results as of 2026-11-17 · retrieved 2026-11-17 · license: CC-BY-4.0