Benchgen

AIME 2026 — Results

RankModelScore
1gpt-5-6-sol99.9
2glm-5-299.2
3gemini-3-1-pro98.3
4beam-1-0-501b-a23b97.7
5inkling97.1
6deepseek-v4-pro96.7
7namazu96.67
8kimi-k2-696.4
9kolibri-196
10kimi-k2-595.8
11qwen3-6-plus95.3
12solar-pro-495.3
13nanbeige4-3-33b95.2
14ripple-65b-preview95
15muse-glimmer94.7
16nemotron-3-ultra-550b-a55b94.2
17qwen3-6-27b94.1
18raon-1-mini-preview93.3
19qwen3-6-35b-a3b92.7
20qwen3-5-397b-a17b91.3
21spark-x2-6-4b91.2
22spark-x2-5-4b90.7
23spark-x2-5-4b90.7
24twil-8b88.3
25maple-preview87.5

AIME 2026 — Evaluations

1 phaseActive

No evaluations yet for this benchmark.