Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematics Reasoning on AIME 2025 (Accuracy and Mean Output Tokens)
Loading...
78.3
Accuracy
Apriel-Reasoner (Ours)
56.46
62.13
67.8
73.47
Apr 2, 2026
Accuracy
Mean Output Tokens
Updated 16d ago
Evaluation Results
Method
Method
Links
Accuracy
Mean Output Tokens
Apriel-Reasoner (Ours)
Size=15B
2026.04
78.3
11,300
Nemotron-Cascade
Size=14B
2026.04
76
19,000
Apriel-Base
Size=15B
2026.04
73.3
16,600
Apriel-Base + RLVR w/ LP
Size=15B, Length Penal...
2026.04
71.7
11,100
Qwen3
Size=14B
2026.04
68
16,900
Phi-4-reasoning
Size=14B
2026.04
57.3
12,500
Feedback
Search any
task
Search any
task