swe-bench evaluation

Anthropic's Claude Opus 4 and Sonnet 4 AI models set new benchmarks in professional environments, with Opus maintaining focus on…