Claude Duet Availability Outside the US: No UK Date or Waitlist Size Announced Yet

Bottom line: According to OrcaRouter’s write-up, Anthropic has not published a UK launch date for Claude Duet, and the waitlist page does not state how many people are queued. Interest data cited in the same piece ranks the US fifth, at an index of 38.

The query is asking for a date the page does not have. OrcaRouter on a page the search record dates to October 23, bylined Rowan Sterling, says “Claude duet uk” is a Breakout query, and that the honest answer is short: Anthropic has published no UK availability, no UK date, and no waitlist that promises either. The same piece says the product page describes Claude Duet as a research preview on a waitlist for Claude Pro and Max subscribers in the United States, powered by Claude Opus 5.5. Three guessed news paths on anthropic.com and claude.com returned 404s.

OrcaRouter’s own split is the first thing to keep. They host Claude Opus 5.5 through an API. They say they have nothing to do with the preview, the waitlist, or the standalone app, and that the article keeps those two things apart from the first sentence. An API key for the model is not a seat in the preview. A seat in the preview is not an API key. If a roundup pastes “we serve Opus 5.5” under “Claude Duet is available,” it has merged the two objects the page asked to keep separate.

We already split a credit floor from a seat cap. The floor was $10,000 and the cap was $50,000. We already had a small-model price that was not the benchmark next to it. Haiku 5.5’s short prompt was ten cents. We already had an announcement that printed a speed and not a score. Mercury 3’s 1,068 was not the scorecard. Duet is the version where the missing object is a queue.

Claude Duet’s queue has no published length

The absence is specific, and it is the useful part.

OrcaRouter says the page does not describe the queue. There is no position, no queue length, no figure for how many people have been invited, and no statement about how long anyone waits. Everything after the submission sits outside what the page documents. A reader looking for a number to watch will not find one, because the number has not been published. That is four missing integers, not one vague “coming soon.” Position would be a rank. Length would be a headcount of people still waiting. Invited would be a headcount of people let in. Wait time would be a duration. A screenshot of a confirmation email has none of those unless the email prints them, and this writeup does not say the email prints them.

The waitlist is also not a sign-up in the ordinary sense, in their telling. Claude Duet exists as a standalone app for iOS and Android and on the web at claude.ai on a computer. Access runs through the preview rather than through a download. You cannot install your way in. The app stores and the URL can be true at the same time as a closed preview. A store listing is a file. A preview is a permission. Installing the file does not, on this account, grant the permission.

According to OrcaRouter, the waitlist is for Claude Pro and Max subscribers in the United States. That is two plan names and one country. It is not “Claude subscribers.” A help-desk answer that says “any paid Claude plan” has widened the gate past the two names on the page.

The country condition is narrower still, and the mechanism is unpublished. The preview is for subscribers in the US. The page, via OrcaRouter, does not describe how the country is established. It does not say whether the waitlist form rejects a submission from outside the United States, what checks run on an account, or whether a plan’s billing country is the thing that counts. Nothing on the page describes changing the country attached to an account, and nothing describes a way to move up or around the queue. An unpublished mechanism is not a hint. It is a missing sentence. A reader in the UK who treats a Breakout query as evidence of a hidden UK switch has the query and not the switch. OrcaRouter’s line is that Anthropic has published no UK availability and no UK date. That is a claim about publication, not a claim that a date exists somewhere off-page.

Powered by Claude Opus 5.5 is a model sentence, not an access sentence. OrcaRouter serves that model through an API and says the Duet app is a different product. A benchmark of Opus 5.5, or a price for Opus 5.5, does not tell you whether the preview let anyone in this week. If you need a price, it is not this note.

The US index is 38, and it is fifth

The attention numbers are real as an index, and they are easy to overread.

OrcaRouter prints a Google Trends card: China 100, Israel 70, Singapore 55, the United Arab Emirates 53, the United States 38, across 66 regions with low-volume regions included. The country where the preview is actually running ranks fifth for interest in the query. Fifth, on that card, is the United States, and the four above it are named. Sixty-six is the width of the index they say they included, not a count of countries where the app is on sale. Low-volume regions included means the ranking was not trimmed to places with a lot of searches. It does not mean each of the 66 has a published user count.

One hundred is the index peak, not one hundred percent of people, and not one hundred million searches. Trends sets the top region in the chosen window to 100 and scales the others. China’s 100 and the US’s 38 are a ratio inside that window: 38 is 38 percent of the peak region’s index, not 38 percent of Chinese searchers and not 38 percent of Americans. Israel at 70 is higher than the country where the preview runs. Singapore at 55 and the UAE at 53 are also higher. None of those four, in the October 23 piece, is described as a place where the preview is open. Interest and availability are already split before you touch the name collision.

The name collision is the part that should shrink the card. Trends also surfaces a rising topic called “Claude Paul Duet,” a person whose name collides with the product query. Some of the volume is not about the product at all. The figures describe search interest in the query, not demand for the app, and they should not be read as a count of people who want access. That caution is in the same article as the sentence “whatever the waitlist is doing, it is not absorbing the attention.” The caution undercuts the sentence. If some of the volume is a person, then “the waitlist is not absorbing the attention” uses a numerator that is not the waitlist. A product query that shares tokens with a person’s name will rank places where that person is searched. China at 100 may be the product, the person, or a mix. The page does not split the mix.

Breakout, on “Claude duet uk,” is a Trends label for a query whose growth is sharp relative to its own past, often because the past was near zero. Breakout is not a headcount of UK users. It is compatible with a small absolute volume. A new two-word query can be Breakout on Tuesday and still be a few hundred searches. OrcaRouter does not print the absolute UK volume. The Breakout tag and the missing UK date can sit together: people are searching a string Anthropic has not, on this account, answered with a date.

What remains unconfirmed

A 404 on a guessed slug is not evidence that Anthropic never published a page. It is evidence that those three slugs were wrong on October 24. The product-page paraphrase in OrcaRouter, the Pro and Max limit, the US limit, the Opus 5.5 line, remains their paraphrase. If Anthropic’s page says something narrower or wider, this note does not catch it.

Paste this, if you are answering the UK query tonight. Anthropic, as of the October 23 writeup, had published no UK availability, no UK date, and no waitlist that promises either. The preview, as described there, is a waitlist for Claude Pro and Max subscribers in the United States. The app is on iOS, Android, and claude.ai, and access is through the preview, not through the install. The queue has no published position, length, invite count, or wait time. The Trends card they printed is China 100, Israel 70, Singapore 55, UAE 53, US 38, and the US figure is an index, contaminated by a person named Claude Paul Duet, not a demand census.

Do not paste an API price for Opus 5.5 as the price of Duet. Do not paste a VPN step. Do not paste “fifth in the world, so the waitlist is failing.” Fifth is a rank on a query index. The waitlist’s size was not published, so it cannot be failing a target that was also not published.

The Breakout query wants a UK date. The page they read does not have one. The queue wants a length. That length is not in the writeup, because they say it is not on the page. Those are the two empty fields. Everything else in the card is a different object.

What this means for you

If you are outside the US, the safest assumption is that there is no confirmed date yet. Join the official waitlist if you need access, and treat third-party timelines as unconfirmed until Anthropic’s own newsroom announces regional availability.