Revolut Changes 2026: Airport Lounges In, WeWork and Perplexity Out

Within a few weeks in the summer of 2026, Revolut announced it was building its own airport lounge in Copenhagen, cut the number of WeWork locations available to its highest-paying subscribers, and replaced a €20-a-month AI subscription with an €8 one.

Taken together they point to a company whose paid plans are being actively reshaped rather than simply expanded. What follows is what actually changed, what is confirmed, and what remains speculation.

Revolut is building an airport lounge in Copenhagen

On 13 August 2026, Revolut announced its first branded airport lounge at Copenhagen Airport, scheduled to open in 2027. Plaza Premium Group comes in as operating and investment partner, meaning Revolut supplies the brand and the customer relationship while an established operator handles hospitality, staffing and catering.

Revolut says the site will be the largest common-use lounge in the Schengen area, and describes Copenhagen as a design and operating blueprint to scale across other European markets from 2027 onwards. Hadi Nasrallah, who runs Revolut's "New Bets" division, characterised the project as unique and gutsy. It follows the announcement of Revolut's first physical store, in Barcelona.

Access terms have not been published. No plan tier — Standard, Plus, Premium, Metal or Ultra — has been confirmed as receiving free or discounted entry.

The strategic logic is familiar from the card industry. American Express built the Centurion network around premium-card loyalty, and Capital One and JPMorgan Chase followed. Owning the space removes a third-party aggregator from the transaction and gives the issuer control over the experience and its cost.

The WeWork benefit shrank on 1 August

This is the change with the clearest financial impact on subscribers.

From 1 August 2026, the WeWork "on demand" network available to Metal and Ultra customers was reduced to roughly 265 locations globally. More than half of those sites are in the United States, where Revolut's retail base is considerably smaller than in Europe.

Cities affected include Paris, Brussels, Milan, London, Edinburgh, Dublin and Prague. In Paris, Brussels and Edinburgh, subscribers were left with no available WeWork sites at all. Paris is also where Revolut is preparing to open its Western European headquarters.

The cause, first reported by the Financial Times, was WeWork seeking higher fees on the back of increased occupancy and demand. Revolut renegotiated the geographic scope rather than withdraw the benefit entirely or pass the increase on to subscribers. Customers were not informed of the change.

The scale of what was lost depends on location. Ultra costs around £55 a month and includes three one-day WeWork passes; Metal is £14.99 a month and includes one. In cities where day passes run £35–£50, three passes can approach or exceed the monthly cost of the Ultra plan on their own.

Perplexity Pro is being replaced by ChatGPT Go

Revolut announced the ChatGPT Go benefit on 30 July 2026 with a formal newsroom post. The consequence arrived separately, by email to Premium, Metal and Ultra customers: from 2 September 2026, Perplexity Pro is removed from plan benefits. One replaces the other.

Revolut's framing emphasised that it was delivering a service used by close to a billion people weekly at no extra cost. On retail price alone, the substitution is a reduction of roughly 60%.

Benefit periods run 12 months for Metal and Ultra, six months for Premium and Revolut Pro, and three months for Plus and Standard. Activation requires a payment method, and according to Revolut's own help pages customers roll onto a paid ChatGPT plan when the benefit period ends unless auto-renewal is switched off beforehand.

That structure is worth understanding, because it applies across most of the bundle. Partnerships of this kind function as customer acquisition channels: the software company reaches a verified, solvent audience without ad spend, absorbing product margin instead, and retains the upsell when the free period lapses.

How connected are these changes, really?

It is tempting to read all three as a single coordinated squeeze. The evidence supports a weaker version of that claim.

What is documented: the WeWork reduction was triggered externally. WeWork raised its rates; Revolut chose to protect the headline plan price and cut geographic coverage instead. That is a decision about where to absorb a cost increase, not a decision to strip value for its own sake.

What is inference: the Perplexity-to-ChatGPT swap has no stated cost rationale, but the price differential makes a margin motive the obvious reading. Revolut has not confirmed one.

What is speculation: that the lounge programme is intended to displace third-party lounge partners such as LoungeKey or DragonPass over time. It is a coherent trajectory — build a network, then substitute owned access for bought access — and it mirrors what has already happened with WeWork. But Revolut has said nothing of the kind, and a handful of European lounges would not replace a global network for years, if ever.

What ties the three together is not a plan so much as a shared constraint. Every included subscription is a cost line inside a business whose valuation depends on the subscription line growing and its margin improving. When any of those cost lines moves, something gives.

The numbers behind the pressure

Revolut's FY2025 results, published on 24 March 2026:

• Group revenue $6.0bn (£4.5bn), up 46%

• Profit before tax $2.3bn (£1.7bn), up 57%, at a 38% margin

• Net profit $1.7bn (£1.3bn)

• Retail customers 68.3 million, up 30%

• Subscriptions turnover $936m (£708m), up 67%

• Paid plan adoption up 42% year on year

Subscriptions were the fastest-growing revenue line in the business.

Against that, Revolut's July 2026 secondary share sale priced shares at $2,017 and valued the company at $115 billion, up from $75bn in November 2025 and $45bn in 2024. That is roughly 68 times net profit. Conventional European banks trade in the mid-teens on earnings.

The gap is the point. Revolut is being priced as a technology platform with a compounding subscription engine, not as a bank — which means the subscription engine has to keep compounding, and its margin has to keep improving, for the valuation to hold.

Notably, none of these changes appear to be costing Revolut customers. Retail growth of 30% ran alongside every one of them.

The incentive structure at the top

Nik Storonsky holds roughly 29% of Revolut, worth over $36bn at the current valuation. His existing package releases shares in stages as the company clears pre-agreed valuation thresholds — a structure closer to Elon Musk's at Tesla than to conventional European executive pay. Storonsky has said it would take his stake to around 40% at a $200 billion valuation, the figure Revolut has signaled to investors for a potential IPO no earlier than 2028.

In August 2026, the Financial Times reported he was in talks over a further award triggering at approximately $500 billion. No terms have been finalised.

This does not explain any individual benefit change, and it would be a stretch to draw a line from a founder's pay package to a specific WeWork contract. It does explain why growth and margin expansion sit at the centre of the company's priorities, and why the subscription business is the part under the most sustained pressure to perform.

Is Revolut Ultra still worth it?

For many subscribers, yes. Even after the WeWork reduction and the Perplexity removal, no comparable UK or EU bundle matches Ultra on breadth. Revolut puts the combined value of the Ultra bundle at around €4,100 a year, though that figure counts every included subscription at full retail price, which few customers would otherwise buy.

The more useful question is how the plan holds up under attrition. Five practical points:

1. Don't subscribe for a single benefit. A plan carried by one perk is one renegotiation away from being worthless — which is precisely what happened to subscribers using WeWork.

2. Value the plan on what would remain if two benefits disappeared. Given the past twelve months, that is now a reasonable base case rather than a pessimistic one.

3. Check the break fee before committing to an annual term. Whether a benefit change permits a fee-free exit is decided case by case.

4. Diarise every free AI or lifestyle trial on the day it is activated. These convert to paid subscriptions by default.

5. Treat the Standard account as a funnel, not a current account. In the UK it compares poorly to mainstream alternatives on its own merits.

Where this leaves the maths

Ultra at £55 a month is £660 a year. Metal at £14.99 is £180. Those are lifestyle purchases, and there is nothing wrong with paying for perks that get used — but they are worth re-pricing whenever the contents change, which is now happening more than once a year.

For anyone concluding that the benefits no longer justify the fee, the alternative is straightforward: downgrade, and route the difference somewhere that pays interest rather than charges a subscription.