Anthropic Releases Claude Opus 4.7 With Major Upgrades

Anthropic has launched Claude Opus 4.7, featuring notable improvements in coding and visual reasoning. The model scored 64.3% on the SWE-Bench Pro benchmark, nearly 10% higher than its predecessor, and can process images at greater resolution while generating better visual assets like UI designs. The release includes a cyberattack detection mechanism and a new Cyber Verification Program to ease restrictions for security professionals. Anthropic also introduced API updates including a new xhigh effort tier, task budgets for cost control, and Claude Code enhancements such as an ultrareview command and auto mode for long-running tasks.