"Claude 3.7 Sonnet is roughly 3.3x more expensive than o3-mini for tokens."
"Claude 3.7 Sonnet excels in reasoning with 84.8% accuracy in GPQA."
"Claude 3.7 Sonnet achieves 96.2% in MATH benchmarks."
"O3-Mini-High outperforms Claude 3.7 Sonnet in identifying critical code issues."