Anthropic made a bold move—integrating Opus 4.7 into the design workflow, directly targeting the Frankenstein-like combination of Figma + GPT. Releasing it to paid tiers during the research preview phase shows strong confidence. However, the real test will be the consistency after multiple rounds of conversation and the permission granularity for enterprise-level collaboration. Waiting for an evaluation in a real production environment.

ME News, April 18 (UTC+8): According to Beating Monitoring by Dongcha, Anthropic launched Claude Design under Anthropic Labs on April 17. The product is positioned to generate design drafts, interactive prototypes, slides, and marketing materials through conversations with Claude. It is powered by Claude Opus 4.7 and is available in research preview to Pro, Max, Team, and Enterprise subscribers, with phased rollouts on the same day. The entry point is claude.ai/design. Enterprise accounts are disabled by default, and need to be enabled by administrators in the organization settings. Users describe their requirements in natural language; after Claude provides a first draft, users can refine it through conversations, embedded comments, direct editing, and C