Anthropic readies Claude Opus 4.7
Anthropic is reportedly launching Claude Opus 4.7 this week alongside a prompt‑based design tool for websites and presentations, while Claude Mythos continues testing for cybersecurity uses. The product signals Anthropic's push into specialised tooling layered on its conversational models. (x.com)
Anthropic is preparing to release Claude Opus 4.7 and a new tool that can build websites and presentations from text prompts as soon as this week. (theinformation.com) The report came from The Information on April 15, 2026, citing a person with knowledge of the products. Anthropic has not posted a public launch announcement for Opus 4.7 on its website as of April 15. (theinformation.com) (anthropic.com) Anthropic’s current top public model is Claude Opus 4.6, announced on February 5, 2026. The company said that version added “adaptive thinking,” new effort controls, and a research preview of Claude in PowerPoint. (anthropic.com) The design tool fits work Anthropic was already moving toward inside office software. In February, Anthropic said Claude in Excel had been upgraded and Claude in PowerPoint had entered research preview. (anthropic.com) A prompt-based design tool works like this: a user types a request in plain English, and the system turns it into layouts, slides, or web pages. The reported product would push Claude beyond chat and coding into software that produces finished visual work. (theinformation.com) (anthropic.com) Anthropic has been steering Claude toward longer, more structured tasks for nearly a year. When it introduced Claude 4 on May 22, 2025, the company said Opus 4 was built for coding, advanced reasoning, and “long-running tasks and agent workflows.” (anthropic.com) The company’s own Opus 4.6 page says the model is aimed at professional software engineering, complex agentic workflows, and high-stakes enterprise tasks. Anthropic also says Opus 4.6 is available on Claude, its application programming interface, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry. (anthropic.com) Anthropic is also expanding its public safety and governance material as it ships stronger models. Its Transparency Hub, updated in February 2026, says the company publishes model reports with capability, safety evaluation, and deployment information for Claude systems. (anthropic.com) Investors treated the report as a threat to incumbent design software. News coverage following The Information’s report said shares of Figma, Adobe, Wix, and GoDaddy fell on April 15 after the story circulated. (thetechportal.com) (finance.biggo.com) If Anthropic ships on that timetable, the next test is whether users treat Claude as a chatbot with better answers or as a tool that can draft the page, build the deck, and hand over something ready to edit. (theinformation.com)