List your company

Mathematical Reasoning on MATH 500 (Accuracy and Classification Metrics)

98.5AccuracyDeepSeek-V3
30.27647.98865.783.412Apr 6, 2026

Evaluation results

MethodLinks
DeepSeek-V3Model=DeepSeek-V32026.0498.556.71.428.797.199.7
Qwen3Model=Qwen32026.049851.6017.592.196.9
Llama4-MaverickModel=Llama4-Maverick2026.0490.383.377.981.280.393.6
Grok-4Model=Grok-42026.0478.776.551.573.385.495.7
Tulu3.1-8BModel=Tulu3.1-8B2026.0448.878.260.687.174.390.9
Llama3.1-8BModel=Llama3.1-8B2026.0446.576.461.880.375.687.7
OLMo2-7BModel=OLMo2-7B2026.0432.976.57190.762.282.1