正文 Markdown
Claude Opus 4.6 is now out! At Box, we've been testing the new model with Box AI on complex areas of knowledge work across many key industries like Financial Services, Life Sciences, and Legal. Overall, Opus 4.6 represents a 10% jump over Opus 4.5 on our hardest knowledge work tasks. In our testing, we have the model perform a variety of complex, real-world tasks on enterprise content in a single shot. This work resembles the sophisticated multi-step work that a professional does in each major industry, and we saw major jumps in accuracy across nearly all areas. What's clear is that Opus 4.6 represents another increase in advanced reasoning, logic, agentic tool use, coding, and more. All of these capabilities are necessary to be able to deploy AI agents at scale in an enterprise. Box customers can build AI agents with Claude Opus 4.6 within the Box AI Studio. Learn more at: https://t.co/9w9MMathqT