Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
General Reasoning on AIME 2024 (Accuracy)
Loading...
89.7
Accuracy (AIME 2024)
TeamTR (Heterogeneous)
4.836
26.868
48.9
70.932
May 1, 2026
Accuracy (AIME 2024)
Updated 2mo ago
Evaluation Results
Method
Method
Links
Accuracy (AIME 2024)
TeamTR (Heterogeneous)
Size=1.7B+8B+14B
2026.05
89.7
TeamTR (Homogeneous)
Size=3×8B
2026.05
88.1
Qwen3 (thinking)
Size=32B
2026.05
81.4
Qwen3-A3B
Size=30B
2026.05
80.4
o4-mini
Size=–
2026.05
79.6
QwQ
Size=32B
2026.05
79.5
Debate + Judge
Size=3×8B
2026.05
71.1
DeepSeek-R1 Distill
Size=70B
2026.05
70
Role-play
Size=3×8B
2026.05
69.9
Debate
Size=3×8B
2026.05
68.7
DAPO
Size=8B
2026.05
41.3
GRPO
Size=8B
2026.05
39.1
GPT-4o-mini
Size=–
2026.05
8.1
Feedback
Search any
task
Search any
task