Google tops OpenAI in math reasoning — 9 to 1

A new mathematical reasoning benchmark has placed Google ahead of OpenAI by an impressive 9-to-1 margin. Google's models demonstrated superior capability in solving complex mathematical problems, including geometry, calculus, and formal proofs. The result marks a significant shift in the AI competitive landscape, where OpenAI traditionally led in reasoning benchmarks. Experts suggest Google's approach — combining multiple specialized models with advanced chain-of-thought techniques — may be driving this superior performance.