Epoch AI - Frontiermath Tiers 1 3 V2: leaderboard
The three lower tiers of Epoch's FrontierMath v2, spanning advanced undergraduate mathematics up through early-career research problems. Every question is original and vetted by working mathematicians, and this is the wide end of the board — a hundred-odd models, and the frontier now clears 90.
Metric: Score. Source: epoch.ai. 120 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | gpt-6.1-sol_max | 93.68 | |
| 2 | gpt-6-astra_max | 93.68 | |
| 3 | claude-opus-5-5_max | 91.23 | |
| 4 | claude-fable-5-1_max | 90.18 | |
| 5 | gpt-6-sol_max | 89.82 | |
| 6 | gpt-5.6-sol_max | 89.12 | |
| 7 | claude-sonnet-5-5_max | 88.77 | |
| 8 | gpt-5.5-pro_xhigh | 87.72 | |
| 9 | claude-fable-5_max | 87.02 | |
| 10 | gemini-4-argon_high | 85.96 |
Interactive version: theaggregate.ai/benchmark?slug=epoch-ai-frontiermath-tiers-1-3-v2 · How It Works · Data refreshed daily, snapshot 2026-10-20.