Benchmark
AIME 2026
All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with…
- Modality
- text
- Categories
- math, reasoning
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 21
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| DiffusionGemma 26B-A4B | 0.691 | 2026-06-10 | evidence | |
| Gemma 4 12B | 0.775 | 2026-05-23 | evidence | |
| Gemma 4 26B-A4B | 0.883 | 2026-04-02 | evidence | |
| Gemma 4 31B | 0.892 | 2026-04-02 | evidence | |
| Gemma 4 E2B | 0.375 | 2026-04-02 | evidence | |
| Gemma 4 E4B | 0.425 | 2026-04-02 | evidence | |
| GLM-5.1 | Z.ai | 0.953 | 2026-04-07 | evidence |
| GLM-5.2 | Z.ai | 0.992 | 2026-06-16 | evidence |
| Inkling-Small | Thinking Machines Lab | 0.955 | 2026-07-30 | evidence |
| Kimi K2.6 | Moonshot AI | 0.964 | 2026-04-20 | evidence |
| MAI-Code-1-Flash | Microsoft | 0.925 | 2026-06-02 | evidence |
| MAI-Thinking-1 | Microsoft | 0.945 | 2026-06-02 | evidence |
| Muse Glimmer-30B | Meta | 0.947 | 2026-08-10 | evidence |
| Qwen3.5-397B-A17B | Qwen | 0.913 | 2026-02-16 | evidence |
| Qwen3.6 Plus | Qwen | 0.953 | 2026-04-02 | evidence |
| Qwen3.6-27B | Qwen | 0.941 | 2026-04-21 | evidence |
| Qwen3.6-35B-A3B | Qwen | 0.927 | 2026-04-16 | evidence |
| Sakana Namazu | Sakana AI | 0.9667 | 2026-08-03 | evidence |
| Seed 2.0 Lite | ByteDance | 0.883 | 2026-02-14 | evidence |
| Seed 2.0 Pro | ByteDance | 0.942 | 2026-02-14 | evidence |
| Solar Pro 4 | Upstage | 0.953 | 2026-08-06 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.