OTIS Mock AIME 2024-25: leaderboard
Mock AIME competition problems from the OTIS math olympiad program. Tests creative mathematical problem-solving at olympiad level. Evaluated by Epoch AI.
Metric: Accuracy (%). Source: epoch.ai. Status: saturated. 288 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 (xHigh) | 100 |
| 2 | GPT-5.6 Sol (Max) | 100 |
| 3 | Claude Fable 5.1 (Max) | 100 |
| 4 | GPT-6 (Max) | 100 |
| 5 | Best Expert | 100 |
| 6 | Claude Fable 5 (High) | 100 |
| 7 | GPT-5.5 Pro (xHigh) | 100 |
| 8 | Claude Fable 5 (Max) | 99.72 |
| 9 | GPT-5.6 Terra (Max) | 99.72 |
| 10 | Qwen 3.8 Max (xHigh) | 99.44 |
| 11 | Grok 4.6 (xHigh) | 99.17 |
| 12 | Claude Opus 5 (Max) | 98.89 |
| 13 | DeepSeek V4 Pro (0813) (Max) | 98.61 |
| 14 | Claude Opus 4.8 (Max) | 98.33 |
| 15 | GPT-5.6 Luna (Max) | 98.33 |
Interactive version: theaggregate.ai/benchmark?slug=otis-mock-aime-2024-25 · How It Works · Data refreshed daily, snapshot 2026-09-05.