ZGCM-1: A Fully Open and Extremely Efficient Foundation Model for Math and Agentic Search

Benchmark Model Rank Results
mathematical-reasoning-on-olympiadbenchZGCM-1#2Accuracy: 76.06
mathematical-reasoning-on-math-500ZGCM-1#14Accuracy: 97.13
mathematical-reasoning-on-aime24ZGCM-1#27Acc: 80.62
mathematical-reasoning-on-aime-2026ZGCM-1#33Accuracy: 75.0
mathematical-reasoning-on-aime-2025ZGCM-1#50Accuracy: 73.33
code-generation-on-livecodebench-v6ZGCM-1#70Pass@1: 0.4686