Problems from the 2025 American Invitational Mathematics Examination used to evaluate advanced mathematical reasoning.
Browse the latest scores, model modes, release dates, and parameter sizes for AIME2025.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
2 parallel-mode results hidden
| Rank | Model | License | |||
|---|---|---|---|---|---|
![]() GPT-5-Pro Thinking EnabledTools | 100.00 | 2025-08-07 | Unknown | Closed | |
![]() Claude Sonnet 4.5 Thinking EnabledTools | 100.00 | 2025-09-30 | Unknown | Closed | |
![]() GPT-5.2 Thinking Level · Extra High | 100.00 | 2025-12-11 | Unknown | Closed | |
4 | ![]() Step 3.5 Flash Thinking EnabledTools | 99.80 | 2026-02-02 | 196B | Free Commercial |
5 | ![]() Claude Opus 4.6 Extended Thinking | 99.79 | 2026-02-05 | Unknown | Closed |
6 | ![]() Gemini 3.0 Flash Thinking EnabledTools | 99.70 | 2025-12-17 | Unknown | Closed |
7 | ![]() GPT-5 Thinking EnabledTools | 99.60 | 2025-08-07 | Unknown | Closed |
8 | ![]() OpenAI o4 - mini Thinking EnabledTools | 99.50 | 2025-04-16 | Unknown | Closed |
9 | ![]() Gemini 2.5 Deep Think Deep Thinking Mode | 99.20 | 2025-08-01 | Unknown | Closed |
10 | ![]() Kimi K2 Thinking Thinking EnabledTools | 99.10 | 2025-11-06 | 1040B | Free Commercial |
11 | Grok 4 Thinking EnabledTools | 98.80 | 2025-07-10 | Unknown | Closed |
12 | ![]() GPT OSS 20B Thinking EnabledTools | 98.70 | 2025-08-06 | 21B | Free Commercial |
13 | ![]() GLM-4.6 Thinking EnabledTools | 98.60 | 2025-09-30 | 355B | Free Commercial |
14 | ![]() GLM-4.6 Thinking Enabled | 98.60 | 2025-09-30 | 355B | Free Commercial |
15 | ![]() GPT OSS 120B Thinking EnabledTools | 97.90 | 2025-08-06 | 11.7B | Free Commercial |
16 | ![]() Step 3.5 Flash Thinking Enabled | 97.30 | 2026-02-02 | 196B | Free Commercial |
17 | ![]() GPT-5-Pro Thinking Enabled | 96.70 | 2025-08-07 | Unknown | Closed |
18 | ![]() Haiku 4.5 128KTools | 96.30 | 2025-10-15 | Unknown | Closed |
19 | ![]() Kimi K2.5 Thinking Enabled | 96.10 | 2026-01-27 | 1000B | Free Commercial |
20 | ![]() DeepSeek V3.2 Speciale Thinking Enabled | 96.00 | 2025-12-01 | Unknown | Free Commercial |
21 | ![]() GLM-4.7 Thinking Enabled | 95.70 | 2025-12-22 | 358B | Free Commercial |
22 | ![]() Gemini 3.0 Flash Thinking Enabled | 95.20 | 2025-12-17 | Unknown | Closed |
23 | ![]() Gemini 3.0 Pro (Preview 11-2025) Thinking Enabled | 95.00 | 2025-11-18 | Unknown | Closed |
24 | ![]() GPT-5 Thinking Enabled | 94.60 | 2025-08-07 | Unknown | Closed |
25 | ![]() Kimi K2 Thinking Thinking Enabled | 94.50 | 2025-11-06 | 1040B | Free Commercial |
26 | ![]() GPT-5.1 Thinking Level · High | 94.00 | 2025-11-12 | Unknown | Closed |
27 | ![]() GPT-5.1 Thinking Level · High | 94.00 | 2025-11-12 | Unknown | Closed |
28 | ![]() DeepSeek V3.2 Thinking Enabled | 93.10 | 2025-12-01 | 671B | Free Commercial |