Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

AI news & research

News ·

The Verge finds Claude Duet useful but a bit of a nag after a week of testing

The Verge's Allison Johnson spent a week with Claude Duet, Anthropic's work agent for professionals and small business owners, a research preview open to Pro, Max and Ultra subscribers in the US through a waitlist. Connected to email, calendar, Docs and Slack, it linked meeting notes to messages and caught a scheduling squeeze, but sometimes pressed on deadlines that were not real.

Claude Duet is Anthropic's agent for people who do not write code, and the first detailed week-long account of using it has arrived from The Verge. The verdict is mixed in a useful way: Duet was good at spotting things its user had missed across email, calendar and chat, and poor at judging how much they mattered. If you are weighing whether to hand an agent your inbox, that split is the thing to understand.

Some context on what Duet is. Anthropic calls it a research preview, aimed at professionals and small business owners rather than shoppers, and limits it to Pro, Max and Ultra subscribers in the US, with a waitlist. It has its own app, and once you connect your work services it offers suggestions as cards you swipe to accept or dismiss. The reviewer connected work email, calendar, Google Docs and Slack, and found Duet quickly linked a remark in a meeting summary to a direct message, and worked out who her manager was from a shared document.

The best moments came from that cross-referencing. Duet noticed that planned time off left only a couple of working days before a big deadline, proposed a fix, offered to update a shared planning document and drafted a Slack note to her editor. The draft copied her habit of writing in lower case, and was labelled as sent with Claude. The weakest came from tone. When she had told an interviewee a story would probably run on Tuesday, Duet treated that as a hard deadline and pressed her before nine in the morning, missing both the hedge and the fact that her editor was in an earlier time zone.

Duet sits at the friendliest end of a family that runs from Claude Code for developers to Claude Cowork for office tasks. Its rivals are OpenAI's Dots and Meta's Muse, both of which lean more towards errands and shopping. The Verge's experience suggests Duet's strength is as a safety net for your own system, catching discrepancies between sources, rather than a replacement for it. The reviewer also found its repeated offers to start her drafts unwelcome.

The limits are those of one person's week. It is a single hands-on review by a journalist, not a measured test, and Duet is still a preview whose judgement about when to interrupt Anthropic says it is still tuning. The privacy cost is real: to be useful it reads every email, document and message you connect.

Our take: The review shows the trade-off with agents that read your whole inbox: real value in spotting what you missed, at the cost of an assistant that sees everything. Anyone outside the US or on a free plan cannot try it yet.

Our reviews:Claude DuetClaude Cowork

Source: The Verge AI

More stories