Claude Duet Waitlist: What to Use Until You Get In (2026)

Claude Duet is Anthropic's new background agent, and as of October 2026 you can only get it by joining a waitlist: it's a research preview for Claude Pro and Max subscribers in the US, with no date for anyone else. If you want what it's known for, a short brief each morning and an agent that keeps working while your laptop is closed, you can set up the same routine today with scheduled Cloud tasks. The method below works in any agent that can run on a schedule.

What is Claude Duet, and who can get it?

Anthropic opened the Duet preview on October 21, 2026. Per Anthropic's Duet page, it works in its own cloud environment, uses the apps you connect, and messages you when work is done or it needs a decision. You can message it by text or voice in one thread. Each morning it summarizes what needs your attention, drafts next steps and then stays quiet until something needs you. TechCrunch's report says it's built on Claude Opus 5.5.

Who can join is narrow:

• US Claude Pro and Max subscribers, through a waitlist. Team and Enterprise plans aren't included yet.
• No separate price. It shares your plan's usage limits, and background checks count toward them. Pro is $20 a month, or $17 a month billed yearly, and Max starts at $100 a month, according to Claude's pricing page.
• Outside the US: Anthropic's page mentions only the US and gives no date for other countries.

Anthropic also lists its limits plainly. It's a preview, it may misread a thread or misjudge urgency, and it can miss something that arrives between check-ins. It asks before sensitive actions such as sending messages or making payments.

What can you do while you wait for Claude Duet?

Build the routine yourself. Most of the value in a Duet-style agent comes from three habits, and none of them needs a special product:

1. One job per task. A morning brief, a weekly competitor check and a deadline watch are three separate tasks. If one fails, the others still run.
2. A fixed time in your own time zone. "Every day at 07:30, Europe/Berlin" is clearer than "every morning".
3. Read freely, write with approval. Let the agent read your tools, but make it ask before it sends, edits or deletes. That one rule is what makes an agent safe to leave alone.

Here's a brief instruction you can paste into any scheduled agent. Change the parts in brackets:

Every morning, build my daily brief.
1. Read my [Linear issues assigned to me or due this week], my [Notion page "This week"] and my [Airtable "Customers" base, rows changed in the last 24 hours].
2. Search the web for news from the last 24 hours about [competitor A], [competitor B] and [my industry keyword].
3. Write the brief in this order: what needs a decision today (max 3 items, one line on why each), what is at risk this week, and what changed overnight.
4. For the top item, draft the next step (a reply, a ticket, a note) but do not send, create or edit anything. Show it to me for approval.
5. If nothing needs me, say so in one line.

Give it two or three days, then tighten. Most briefs need edits to the "what needs a decision" rule before they feel right, such as which senders or customers always count as urgent.

How does Zemith Cloud compare with Claude Duet?

Zemith Cloud is a smaller, schedule-driven agent that you can use now, in any country Zemith runs in. Duet is the more ambitious product: it watches your apps on its own and talks to you in a thread. Here's the honest side-by-side, as of October 2026:

Duet's app list is from TestingCatalog's launch coverage; Zemith's from its plugin directory in the product. The tasks themselves are identical in spirit: read, research, draft, wait for you.

Doing it in Zemith, step by step

Cloud is part of Zemith Professional ($29.99 a month, $299.88 a year). The Plus plan at $15.99 doesn't include it. Here's the brief above as a schedule:

1. Connect your apps. In Zemith, add the plugins you want the brief to read, such as Linear, Notion and Airtable. Reads run freely. Any write asks for your approval first.
2. Open Cloud in the sidebar and create a Cloud Project called "Daily brief". Paste the instruction into the project's instructions and turn on Enable Cloud Memory, so it remembers what it flagged yesterday.
3. Go to Schedules and choose New schedule. Paste your task into Task instructions.
4. Set Frequency to Daily, Time to 07:30 and your Time zone, and pick the model. Claude 5.5 Opus is the closest match to what Duet runs on; Claude 5.5 Sonnet costs less per run.
5. Choose Create schedule. At 07:30 you get a push notification. If the agent wants to draft a Linear comment or add a Notion note, the task stops with a Needs approval notification until you say yes.

You can add a second, weekly schedule for the things that don't need a daily look, like "Check these three competitors' pricing and changelog pages and save a one-page Word summary". More on what Cloud can do is on the Zemith AI agent page, and Zemith for founders shows how solo operators use it.

What does a daily brief cost in credits?

Each Cloud run draws on the same monthly credits as chat. Claude 5.5 Opus costs 80 credits per 100 input tokens, 400 per 100 output tokens and 4 per 100 cached input tokens, and each web search is 400 credits (Zemith's price list, October 2026). Claude 5.5 Sonnet costs 46, 226 and 5.

Here's an assumed brief, not a measurement: 4 searches, 40,000 new input tokens from your apps and pages, 150,000 cached tokens of project context, and a 3,000-token brief.

• On Claude 5.5 Opus: 1,600 (searches) + 32,000 (input) + 6,000 (cached) + 12,000 (output) = 51,600 credits
• On Claude 5.5 Sonnet: 1,600 + 18,400 + 7,500 + 6,780 = 34,280 credits

Run daily for 30 days, that's about 1.55 million credits on Opus (half of Professional's 3 million) or about 1.03 million on Sonnet. Real runs vary with how much your apps return, and chat and research draw on the same pool. Top-ups start at $10 for 800,000 credits.

Can you get Claude Duet in the UK or Europe?

Not today, as far as Anthropic's published information goes. The Duet page names Pro and Max subscribers in the US and says nothing about other countries. Nobody outside Anthropic knows whether or when that changes, so treat any date you see on a blog as a guess.

If you're in Europe and want a similar agent now, the setup above works there. We covered the same situation for OpenAI's agent in ChatGPT dots in the UK and EU, including an Outlook version of the morning brief.

What doesn't Zemith Cloud do that Claude Duet does?

Quite a bit, and it's worth knowing before you pick:

• It doesn't act on its own initiative. Duet watches your apps and speaks up. Cloud runs the tasks and schedules you set, nothing more.
• No conversation thread with the agent. You send follow-up instructions to a running task, but you can't text it from a messaging app.
• Fewer apps. There's no Gmail, Slack or Stripe plugin, and no purchases.
• No live browser. Cloud's sandbox is headless: it can fetch pages and run scripts, but there's no browser you watch or take over. Anthropic says Duet doesn't show its browser activity live either.
• Schedules are once, daily or weekly, with up to 35 active, and up to 3 tasks running at once. Each run is capped at five hours.
• No deep research inside Cloud. Cloud tasks search and read the web, but a long cited report is a chat feature.

If your work runs on Gmail and Slack, Duet is the better fit once you're in. If it runs on Microsoft 365, Notion, Linear or Airtable, Cloud covers it today. If you hit the scheduled-task caps elsewhere, see our comparison of ChatGPT's scheduled tasks limit and Cloud.

How we compared

On October 27, 2026, we read Anthropic's Duet page and pricing page and two launch reports for Duet's access, price, model, apps and limits. Zemith's Cloud features, plan access, labels and credit rates come from its product code and pricing the same day. The credit example is arithmetic from published rates for an assumed task. We haven't used Duet, and we didn't test agent quality on either product.

The bottom line

Claude Duet is a promising agent, but it's a US-only waitlist today and nobody has said when it will reach anyone else. You don't have to wait to get a daily brief and background work with approval before anything is written. Zemith Professional at $29.99 a month runs it on a schedule, with Claude 5.5 Opus or any other top model, and it works wherever Zemith does. Start with one brief, tune it for a week, and if Duet reaches you later, you'll already know what you want from it.