Claude Duet: Anthropic launches a proactive AI agent app for professionals and small businesses

Anthropic has opened a research preview of Claude Duet, a new app in which its Claude AI acts as a proactive agent: it comes to you with what needs attention, works on tasks in the background and checks with you before acting on your behalf. The company announced it on Wednesday, Oct. 21, and says the app is built for “professionals and small business owners with more on their plates than they can track.”

The launch puts Anthropic squarely in the crowded race for AI agents. The Verge described Duet as a competitor to Meta’s Muse and OpenAI’s Dots, and TechCrunch called it “Anthropic’s bet on agents.”

What is Claude Duet?

Claude Duet is a standalone app, separate from the main Claude app, available on iOS, Android and the web at claude.ai. It runs on Anthropic’s Claude Opus 5.5 model and carries over the memory, skills and connected apps a user has already set up in Claude, according to Anthropic.

The pitch is that Duet comes to you first. Anthropic says:

• Each morning it sends a short brief on the most important items across your connected apps.
• Through the day it keeps tabs on those apps and messages you only when something needs your attention. It checks connected accounts “every few hours” rather than nonstop, partly to save usage.
• On request it takes on tasks “of all shapes,” from scheduling meetings to building dashboards, financial or sales forecasts, prototypes and web apps.

Duet has its own virtual machine with a browser in the cloud, so it can keep working after you close the app or turn off your phone. You talk to it by typing or by voice in one ongoing thread.

How does the Claude Duet app work with other tools?

Anthropic says Duet can use more than 6,000 connectors and plugins in the Claude Marketplace. Examples the company lists include finding missed Gmail and Slack messages, checking QuickBooks or Stripe for unpaid invoices and drafting reminders, preparing payroll in Gusto, drafting job posts for LinkedIn, sending agreements through Docusign and managing a Shopify store.

It can also buy things for a business. Purchases go through Stripe’s Link digital wallet or Shopify’s Shop Pay, and each one uses a one-time virtual card, so the real card number isn’t shared, according to Anthropic’s help center.

Who can get Claude Duet, and what does it cost?

For now, access is limited:

• US only, for Pro and Max subscribers, through a waitlist. Invitations go out in waves.
• Not yet available on Team or Enterprise plans; Anthropic says it plans to add them later.
• No extra charge during the research preview. Duet is included in Pro and Max, but it shares the plan’s usage limits, and its background work, such as the morning brief and check-ins, counts toward them.

The help center adds that people admitted to the preview get one full usage-limit reset, available until Nov. 30, 2026.

What Claude Duet can’t do (yet)

Anthropic is unusually direct about the limits. It calls this “an early version” that “still might misread a thread or miscalibrate what’s urgent.” Among the restrictions the company lists:

• Before it sends a message, makes a payment, signs a document or deletes anything, Duet shows exactly what it is about to do and waits for approval.
• Some actions are off-limits even with approval, including trading securities and moving money between accounts.
• It isn’t as strong at personal errands such as booking tickets, and you can’t watch or take over its browser.
• It doesn’t make phone calls or send texts to other people, and Microsoft 365 isn’t supported yet.
• You can’t choose a model; it runs on Claude Opus 5.5.

Is Claude Duet safe to connect to email and payments?

Anthropic says it has layered several safeguards on top of the approval step. Duet never directly sees passwords or sign-in tokens, the company says, and its actions are screened by the same “auto mode” classifier used for Claude Code. It also launched with a new classifier and safety prompt that check each message and are designed to reject requests to share the most sensitive information, such as credit card numbers and government ID. TechCrunch noted that by default the agent turns down such requests.

Anthropic also acknowledges the risk of prompt injection, where text in an email or web page tries to trick the agent. Its help center says “no defense is perfect,” which is one reason consequential actions require the user’s sign-off.

FAQ

When was Claude Duet announced? Anthropic announced the research preview on Oct. 21, 2026.

Is Claude Duet free? It is included at no extra cost for US Pro and Max subscribers during the preview, but it uses the plan’s usage limits.

Is it part of the regular Claude app? No. For now it is a separate app; Anthropic says early feedback will shape how the two work together.