Benchmark
MathArena Putnam 2025
Math benchmarks measure how well models solve maths problems. Scores are percentages.
All math benchmarksScores from MathArena data.
- Models measured
- 4
- Scores recorded
- 4
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 65.2%
- Mean
- 75.8%
- Median
- 12.1%
- Std dev
- 48.3%
- 10th percentile
- 77.5%
- 90th percentile
- 48.3%
- Min
Leaderboard
| # | Model | |
|---|---|---|
| 1 | GPT-5 Pro· MathArena data | 77.5% |
| 2 | Gemini 3 Pro Previewpreview· MathArena data | 75.8% |
| 3 | DeepSeek V3.2 Speciale· MathArena data | 59.2% |
| 4 | GPT 5.1high· MathArena data | 48.3% |
Tokens the model generated per task, as measured by the source. Comparable within this benchmark only.
Data credits: MathArena (ETH SRI / INSAIT), MIT — cite dekoninck2026matharena (licence)