List your company

Mathematical Reasoning on Math_500 CodaSet OOD (test)

90.33Accuracy (%)DeepSeek-V3.2-Exp
75.114879.064983.01586.9651May 25, 2026

Evaluation results

MethodLinks
DeepSeek-V3.2-Exp2026.0590.333.5
DeepSeek-V3.12026.0589.865.6
Qwen3-235B2026.0585.985.3
DecoRBase Model=DeepSeek-V3.12026.0585.513.1
gpt-oss-120b2026.0583.181.9
EmbedLLM2026.0582.715.1
MODEL-SAT2026.0581.982.9
Kimi-K2-Instruct2026.0581.3115.2
RouterDC2026.0580.643.9
DecoRBase Model=gpt-oss-120b2026.0580.371.4
Random Router2026.0580.184
gemma-3-27b-it2026.0578.041.3
LLM Router2026.0577.571.2
KNN Router2026.0577.11.1
Mistral-3.2-24B2026.0576.641.4
gemma-3-12b-it2026.0575.71