> don't feel that short vs long answers make any difference The “thinking” models are really verbose output models that summarise the thinking at the end. Thes

> don't feel that short vs long answers make any difference The “thinking” models are really verbose output models that summarise the thinking at the end. These tend to outperform non-thinking models, but at a higher cost. Anthropic lets you see some/all of the thinking so you can see how the model arrived at the answer.