AIME 2024
Harbor adapterA benchmark for evaluating AI's ability to solve challenging mathematics problems from the 2024 AIME - a prestigious high school mathematics competition.
Auto57%Math
- Samples
- 30
- Released
- N/A
- Scores
- 0
How to run
ltb · aime-2025
$ npm i -g lettokenburn-cli$ ltb bench pull aime-2025 --n 50$ ltb bench run ./aime-2025 -a claude-code -m anthropic/claude-opus-4-8$ ltb bench push
Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs
Scores
12 sourced entries| Harness | Source | Confidence | Link | |||
|---|---|---|---|---|---|---|
| Claude Opus 4.6 | 100.0 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| Gemini 3.1 Pro | 100.0 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| GPT-oss 120B | 97.9 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| Kimi K2.5 | 96.1 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| Gemini 3 Pro | 96.0 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| Grok 3 | 93.3 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| DeepSeek V3.2 | 89.3 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| Mistral Large | 88.0 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| DeepSeek R1 | 87.5 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| MiniMax M2.5 | 86.3 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| GLM-5 | 84.0 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
| Claude Sonnet 4.6 | 52.8 | N/A | Self-reported | low | Jun 1, 2026 | kaggle.com |
Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.
Reproduced by the community
All runsLoading community runs…
About
AIME 2025 (American Invitational Math Exam) — 30道AIME级数学题. Scoring: 准确率. Difficulty: 极高. Imported from benchmark_research.md (2026-06).