AIME
Can the model solve difficult problems that require more than factual recall?
Reasoning11 ranked models12 reported values2 reportsHigher is better
Top rankings
One row per model, using its best reported score across effort settings.
| Rank | Model | Best score | Best reported setting |
|---|---|---|---|
| 1 | GLM-5.2Z.ai | 99.2% | max effortZ.ai report1 value · 1 reportJun 16, 2026 · source |
| 2 | GPT-5.5OpenAI | 98.3% | Reported configurationZ.ai report1 value · 1 reportJun 16, 2026 · source |
| 3 | Gemini 3.1 ProGoogle | 98.2% | Reported configurationZ.ai report2 values · 2 reportsJun 16, 2026 · source |
| 4 | Qwen3.7-MaxQwen | 97.0% | Reported configurationZ.ai report1 value · 1 reportJun 16, 2026 · source |
| 5 | Kimi K3Moonshot AI | 96.7% | Reported configurationMoonshot AI report1 value · 1 reportApr 20, 2026 · source |
| 6 | Claude Opus 4.8Anthropic | 95.7% | Reported configurationZ.ai report1 value · 1 reportJun 16, 2026 · source |
| 7 | GLM-5.1Z.ai | 95.3% | Reported configurationZ.ai report1 value · 1 reportJun 16, 2026 · source |
| 8 | GPT-5.4OpenAI | 95.0% | xhigh effortMoonshot AI report1 value · 1 reportApr 20, 2026 · source |
| 9 | DeepSeek V4 ProDeepSeek | 94.6% | Reported configurationZ.ai report1 value · 1 reportJun 16, 2026 · source |
| 10 | Kimi K2.6Moonshot AI | 93.3% | thinking effortMoonshot AI report1 value · 1 reportApr 20, 2026 · source |
| 11 | Claude Opus 4.6Anthropic | 92.8% | max effortMoonshot AI report1 value · 1 reportApr 20, 2026 · source |
Effort curve
Every sourced cost-linked effort value for this exact version. Lines connect complete sweeps only.
No cost-linked effort sweep for this version.
Definition and comparison boundarypublic methodology
Can the model solve difficult problems that require more than factual recall?
Multi-step reasoning on academic, mathematical, or abstract problems. Higher is better. The value is the percentage reported in this lab's table.
Harness, tools, prompts, attempt count, and benchmark version can materially change the result. Compare within one reporting context.
Method / source