Opus 4.1

Claude Opus 4.1 builds upon its predecessor, offering significantly improved capabilities in software development, multi-step reasoning, and agent-like task execution. It achieves a remarkable 74.5% accuracy on the SWE‑Bench Verified coding benchmark, surpassing Opus 4 and earlier. The update particularly shines in complex codebases—providing improved debugging, multi-file code refactoring, and high-detail analysis—making it a preferred choice for developers and enterprise