A benchmark based on the 2026 American Invitational Mathematics Examination for evaluating advanced mathematical reasoning.
Browse the latest scores, model modes, release dates, and parameter sizes for AIME 2026.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
| Rank | Model | License | |||
|---|---|---|---|---|---|
 GLM-5.2 Thinking Enabled | 99.20 | 2026-06-13 | 753.3B | Free Commercial | |
Inkling Thinking Enabled | 97.10 | 2026-07-15 | 975B | Free Commercial | |
 Kimi K2.6 Thinking Enabled | 96.40 | 2026-04-20 | 1000B | Free Commercial | |
4 | GLM 5.1 Thinking Enabled | 95.30 | 2026-03-27 | 75.4B | Free Commercial |
5 |  Qwen 3.6 Plus Preview Thinking Enabled | 95.30 | 2026-03-31 | Unknown | Closed |
6 | Qwen3.6-27B Thinking Enabled | 94.10 | 2026-04-22 | 27B | Free Commercial |
7 |  Opus 4.5 Extended Thinking | 93.30 | 2025-11-25 | Unknown | Closed |
8 | GLM-4.7 Thinking Enabled | 92.90 | 2025-12-22 | 358B | Free Commercial |
9 |  DeepSeek V3.2 Thinking Enabled | 92.70 | 2025-12-01 | 671B | Free Commercial |
10 | GLM-5 Thinking Enabled | 92.70 | 2026-02-11 | 744B | Free Commercial |
11 | Qwen3.6-35B-A3B Thinking Enabled | 92.70 | 2026-04-16 | 35B | Free Commercial |
12 | Kimi K2.5 Thinking Enabled | 92.50 | 2026-01-27 | 1000B | Free Commercial |
13 | Qwen3.5-397B-A17B Thinking Enabled | 91.30 | 2026-02-16 | 39.7B | Free Commercial |
14 |  Gemini 3.0 Pro (Preview 11-2025) Thinking Enabled | 90.60 | 2025-11-18 | Unknown | Closed |
15 |  Gemma 4 31B Thinking Enabled | 89.20 | 2026-04-02 | 30.7B | Free Commercial |
16 | Gemma 4 26B A4B Thinking Enabled | 88.30 | 2026-04-02 | 25.2B | Free Commercial |
17 | Gemma 4 E4B Thinking Enabled | 42.50 | 2026-04-02 | 8B | Free Commercial |
18 | Gemma 4 E2B Thinking Enabled | 37.50 | 2026-04-02 | 5.1B | Free Commercial |
