TL;DR: Anthropic has opened the Claude Duet research preview, a standalone app in which Claude works as a proactive agent: it sends a morning brief, checks your connected apps every few hours, runs tasks on its own cloud computer and asks for approval before it sends, pays, signs or deletes anything. Access is by waitlist for US Pro and Max subscribers only, it runs on Claude Opus 5.5, and it shares your plan’s usage limits.
Anthropic says Claude Duet is built for “professionals and small business owners with more on their plates than they can track.” Where the regular Claude app waits for you to type, the Claude Duet research preview is designed to come to you first, flag what needs attention and keep working after you close the app. Here is what the company has announced, who can get in and the limits worth knowing before you join the waitlist.
What is Claude Duet?
According to Anthropic’s announcement, Claude Duet is a new app “where Claude works as a proactive agent. It comes to you with what needs your attention, completes tasks you ask of it, keeps working even when you step away, and checks with you before it acts on your behalf.” It is available as a standalone app on iOS and Android and on claude.ai.
Anthropic says it designed the product around Claude Opus 5.5 and built it “to decide what needs you, what can wait, and when to check with you first.” If you have used Claude Cowork, the difference is the direction of the conversation: Duet is meant to start it.
How a typical day works, per Anthropic
- Morning brief: each morning it sends a short summary of the most important items across your connected apps. The Claude Help Center guide says it looks across your email and calendar and can start first steps such as drafting replies for you to review.
- Periodic check-ins: it checks your connected accounts every few hours rather than nonstop. Anthropic says checking constantly “would use up your plan’s limits much faster.”
- Background tasks: it has a virtual machine with its own browser, so your laptop or phone does not need to stay on. You get a notification when a task finishes or needs review.
- One thread: you message it by typing or voice in one ongoing thread, and you can open separate channels for different topics that share the same memory.
- Bigger jobs: research briefs, financial models, slide decks and dashboards come back as files in a “Make” tab, according to the help article.
Who can get Claude Duet right now
The preview is narrow. Anthropic’s help article lists three requirements: a Pro or Max plan, a Claude account in the US, and an invitation from the waitlist. Team and Enterprise plans are not eligible yet, though Anthropic says it plans to bring Duet to them later.
| Detail | What Anthropic says |
|---|---|
| Status | Research preview, opened October 21, 2026 |
| Plans | Pro and Max (Team and Enterprise later) |
| Region | United States only during the preview |
| Access | Waitlist, invitations sent in waves |
| Platforms | iOS, Android and claude.ai |
| Model | Designed around Claude Opus 5.5 |
| Cost | Included in Pro and Max; shares your plan’s usage limits |
One practical tip from the help article: access is tied to the email address you give the waitlist, so use the same address as your Claude account and the Apple or Google account on your phone.
What Claude Duet can connect to
Duet carries over the memory, skills and connectors you already set up in Claude. Anthropic says it can draw on more than 6,000 connectors and plugins in the Claude Marketplace, and gives examples such as:
- Finding missed Gmail and Slack messages and helping you reply
- Checking Intuit QuickBooks or Stripe for unpaid invoices and drafting reminders
- Preparing payroll in Gusto
- Drafting a job post and, once you approve it, posting it to LinkedIn
- Preparing agreements for signature through Docusign
- Drafting follow-ups from HubSpot deal history and building social campaigns in Canva
- Setting up and managing a Shopify store
For sites without a connector, it uses its own cloud browser. Under the hood many of these integrations rely on the same plumbing we covered in our explainer on MCP in AI.
Anthropic also says Duet can buy things for a business using Link, Stripe’s digital wallet: Duet confirms the item, Link asks you to approve the spend, and payment goes through a virtual card so your real card number is not shared. It can also check out with Shopify’s Shop Pay. The company says sign-in to sites using logins saved in 1Password is “coming soon.”
Safety: what stays in your hands
The core rule is approval before consequential actions. Anthropic says that before 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.” Some things are off-limits even with approval, including trading securities and moving money between accounts.
The safeguards Anthropic lists
- Credential handling follows the design of Claude Managed Agents, so Duet “never directly sees your passwords or sign-in tokens.”
- Actions are checked by Anthropic’s auto mode classifier, the same safeguard used with Claude Code.
- Prompt-injection defenses come from the company’s browser-use research.
- Red teams and an external firm penetration-tested the app, and it is covered by Anthropic’s public bug bounty.
- A new dedicated classifier and safety system prompt are meant to block sharing of the strictest tier of information, such as credit card numbers and government ID. Anthropic says other sensitive data, like personal finances, is passed on only when you explicitly instruct it.
Those claims come from Anthropic and have not been independently tested yet. Agents that read inboxes and act in the background are exactly where prompt injection and misfires cause trouble. OpenAI’s own disclosures, which we covered in OpenAI’s misalignment incident report, show why the approval step matters.
Known limits during the preview
Anthropic is unusually direct about the rough edges. The announcement says Duet “still might misread a thread or miscalibrate what’s urgent,” that personal errands such as booking tickets “aren’t quite as strong,” and that it cannot yet show you its browser. It also lives in a separate app from the main Claude app for now.
Usage is the other catch. Duet’s background checks count toward your plan’s limits, and Anthropic says larger tasks like building a financial model or a deck “require more model turns and background work.” If you already push your Pro limits with chat or coding, expect Duet to compete for the same budget. Our guide to tokens in AI explains why long agent runs eat allowances quickly.
Why the Claude Duet research preview matters
Most assistants still wait for a prompt. Duet is Anthropic’s bet that the useful version of an agent is one that watches your work apps on a schedule, surfaces a short list and does the busywork in the background while you keep the final say.
The approach puts Anthropic in more direct competition with other personal-agent efforts, and it builds on patterns we described in our roundup of agentic AI workflow examples. CNBC also covered the launch on October 21.
Should you join the waitlist?
If you are a US-based Pro or Max subscriber who lives in Gmail, Calendar, Slack and a few business tools, it is worth signing up: it costs nothing extra and invitations arrive in waves. Start by connecting only the apps you are comfortable with it reading, keep the approval prompts on and watch how fast it uses your limits in the first week. Everyone else, including Team and Enterprise customers and users outside the US, will have to wait for Anthropic to widen access.
The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.
