FrontierMath Benchmark Leaderboard
A benchmark of hundreds of original, exceptionally challenging mathematics problems crafted and vetted by expert mathematicians, covering most major branches of modern mathematics from number theory and real analysis to algebraic geometry and category theory.
Leaderboard
Top 13 models on FrontierMath Benchmark Leaderboard (scores from public evaluations).
- 1GPT-5.447.6% on FrontierMath Benchmark Leaderboard
- 2GPT-5.240.3% on FrontierMath Benchmark Leaderboard
- 3GPT-5.5 Pro39.6% on FrontierMath Benchmark Leaderboard
- 4GPT-5.535.4% on FrontierMath Benchmark Leaderboard
- 5GPT-5.126.7% on FrontierMath Benchmark Leaderboard
- 5GPT-5.1 Instant26.7% on FrontierMath Benchmark Leaderboard
- 5GPT-5.1 Thinking26.7% on FrontierMath Benchmark Leaderboard
- 8GPT-526.3% on FrontierMath Benchmark Leaderboard
- 9GPT-5 mini22.1% on FrontierMath Benchmark Leaderboard
- 10o315.8% on FrontierMath Benchmark Leaderboard
- 11GPT-5 nano9.6% on FrontierMath Benchmark Leaderboard
- 12o3-mini9.2% on FrontierMath Benchmark Leaderboard
- 13o15.5% on FrontierMath Benchmark Leaderboard
Models tracked
Models with frontiermath in their evaluation profile.
- No models linked yet.