>Opus 4.7 is smarter than even Gemini 3.1 Pro on nearly every metric,
Outside of coding, claude models are pretty meh. GPT and Gemini are the workhorses of science/math/finance.
And even on coding, they are mostly good at generating new code.
They sure are not at thorough analysis or debugging, etc.
Not in my fields of science: Genetics and neuroscience. The combination of Opus 4.7 Adaptive used with well structure project folders is amazingly useful.