AIME
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
AIME 2024
89.8Pass@16
6
AIME-Extend
52.67Accuracy
6
AIME 25
28.7Maj Score
6
AIME 24
30.8Maj Score
6
AIME25
76.3Accuracy
6
AIME25
83.33Pass@16
6
AIME24/25
5,888Throughput (token/s)
6
AIME26
44.06AIME26 Accuracy
5
AIME 2025
99Score
5
AIME 2026
97.5Score
5
AIME24@32 (test)
73.33Accuracy
5
AIME23@32 (test)
42.22Accuracy
5
AIME22 @32 (test)
46.67Accuracy
5
AIME 2023
13.33Accuracy (AIME 2023)
5
AIME 2025
62.2Accuracy
5
AIME 2025
83.33Pass@12
5
AIME 2024
87.78Pass@12
5
AIME24
91.3Accuracy
5
AIME 2024 + 2025
3.9Avg Diversity
5
AIME 2025
89.47AUROC
5
AIME 2024
87.39AUROC
5
AIME24
50ACC
5
AIME
95.56Pass@1
5
AIME25
63.33Pass@8 Score
5
AIME24
83.33Pass@8
5