IneqMath: leaderboard
Olympiad-level inequality proof benchmark evaluating both final-answer correctness and step-wise reasoning soundness for chat and reasoning LLMs.
Metric: Overall Accuracy (imported). Source: ineqmath.github.io. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 42 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 47 | #98 |
| 2 | O3 Pro | 46 | #116 |
| 3 | O3 | 37 | #124 |
| 4 | GPT-5 Mini | 30.5 | #186 |
| 5 | O4 Mini | 15.5 | #180 |
| 6 | DeepSeek R1 0528 | 9.5 | #222 |
| 7 | O3 Mini | 9.5 | #272 |
| 8 | Kimi K2 | 9 | #243 |
| 9 | O1 | 8 | #183 |
| 10 | DeepSeek V3 (0324) | 7 | #340 |
| 11 | Qwen 3 4B | 5.5 | #829 |
| 12 | Grok 3 | 3.5 | #304 |
| 13 | GPT-4o | 3 | #343 |
| 14 | Gemini 2.0 Flash | 3 | #342 |
| 15 | Qwen 2.5 7B | 3 | #922 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ineqmath · How It Works · Data refreshed daily, snapshot 2026-10-20.