Claude 3.7 Sonnet's best trick is letting you choose when it thinks
Anthropic made Claude 3.7 Sonnet a hybrid reasoning model instead of a separate thinking product. Good. Users do not want a model menu with homework. They want control when the task deserves it.