Claude Duet: Anthropic's Proactive AI Agent, Explained

The short answer

Claude Duet flips the usual chatbot flow. Instead of you opening a chat and typing a prompt, Claude Duet comes to you: it sends a short morning brief across your connected apps, watches for things that need your attention during the day, and keeps working on tasks you hand it even after you step away^[1]. It's built around Claude Opus 5.5^[3] and is included in Pro and Max plans for the research preview^[1].

Availability and pricing

Claude Duet is available as a standalone app on iOS and Android, and on claude.ai^[1]. For now it's a research preview gated by a waitlist, starting with Pro and Max subscribers in the US^[1].

There's no separate fee: Claude Duet is included in Pro and Max and shares the same usage limits as the rest of your Claude plan. Its background checks and larger tasks draw from those existing limits, so heavier projects use more of your plan^[1].

What Claude Duet can do

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 actually needs you^[1]. You can hand it administrative tasks — like scheduling meetings — or larger projects such as building dashboards, creating financial or sales forecasts, prototyping, and building web apps^[1].

Early testers describe the value as catching things you'd miss on your own. Brand strategist Mya Dalton had a client call land right before her flight; Claude Duet saw it on her calendar, prepared a brief, and offered to move it. Broker Jess Provencher got a morning rundown that flagged a property showing booked on the wrong day. And restaurateur Amin Ebrahimi used it for accounting, customer support, and marketing — reconciling invoices store by store^[1].

How it works

Claude Duet runs in a virtual machine with its own browser, so it doesn't need your laptop or phone to stay on. It carries over your memory, skills, and connected apps from your Claude account, and you message it by typing or voice in one ongoing thread^[1]. It checks your connected accounts every few hours rather than constantly, which keeps background monitoring from draining your plan limits too quickly^[1].

Important decisions stay with you. Before Claude Duet sends a message, makes a payment, signs a document, or deletes anything, it shows you exactly what it's about to do and waits for your approval. A few things are off-limits even with approval — like trading securities or moving money between accounts — and you can disconnect any app it has access to^[1].

Connectors and integrations

Claude Duet can use the 6,000+ connectors and plugins in the Claude Marketplace^[4]. Among the highlights Anthropic calls out^[1]:

• Find the Gmail and Slack messages you've missed and help you reply.
• Check Intuit QuickBooks or Stripe for unpaid invoices and draft reminders.
• Prepare and run payroll in Gusto, with hours, bonuses, and reimbursements filled in.
• Draft a job posting and, after you approve, post it to LinkedIn.
• Prepare agreements and send them for review and signature through Docusign.
• Draft follow-up emails from contact and deal history in HubSpot.
• Turn a product launch into a social campaign in Canva, and set up or manage a Shopify store.

It can also make business purchases using Stripe's Link digital wallet and Shop Pay, paying with a virtual card so your real card number isn't shared. 1Password sign-in support is coming soon^[1].

Safety and discretion

Anthropic says Claude Duet never directly sees your passwords or sign-in tokens, reusing the credential design from Claude Managed Agents, plus the auto-mode classifier and browser-use prompt-injection defenses^[1][2]. It ships with a dedicated classifier and safety system prompt that rejects requests to share the strictest information — including credit card numbers and government ID — and only lets other sensitive data (health, personal finances) through when you've explicitly instructed it to send it^[2]. Red teams and an external firm penetration-tested Claude Duet, and it's covered by Anthropic's public bug bounty^[1].

Should you join the waitlist?

Verdict: If you're a US Pro or Max subscriber juggling inbox, calendar, invoicing, and follow-ups, Claude Duet is worth the waitlist — it's the first Anthropic product that acts before you ask. Just set expectations: it's still early, can misread a thread or miscalibrate what's urgent, is weaker at personal errands, and can't yet show you its browser^[1].

Frequently asked questions

Is Claude Duet free?

It's included in Claude Pro and Max plans during the research preview, with no separate fee — but it shares your plan's usage limits^[1].

Does Claude Duet act without asking?

No. It reads, drafts, and prepares work on its own, but it waits for your approval before sending, paying, signing, or deleting anything — and some actions stay off-limits entirely^[1].

Which model powers Claude Duet?

Anthropic designed it around Claude Opus 5.5, its agentic-coding and knowledge-work flagship^[3].