▸ AI/LLM · Anthropic
Anthropic Ships Claude Opus 4.7 With 87.6% SWE-bench Score
Anthropic released Claude Opus 4.7 on April 16, delivering significant improvements in software engineering, vision capabilities (3x higher resolution), and long-running agent tasks. The model scored 87.6% on SWE-bench Verified and introduced new features including xhigh effort level and /ultrareview for code review. Anthropic also launched Claude Design, a new tool for creating visual outputs.
The bottom line: Marketing data teams using AI coding tools should evaluate Opus 4.7 for pipeline development and automation tasks where sustained reasoning over long horizons is required.