Anthropic today released Claude 3.7 Sonnet, introducing hybrid reasoning architecture that allows developers to dynamically specify token budgets for complex logical tasks while maintaining low latency for straightforward completions. The model establishes new state-of-the-art benchmarks in SWE-bench Verified coding tests and complex multi-file engineering workflows.
Original publisher attribution: TechCrunch AI
Xem ấn phẩm gốc