Mullvad Has Not Raised Its Price Since 2009. Revenue Grew 7.7x Anyway.

Quick answer (2026): Mullvad VPN AB's turnover was SEK 40.7M (2020), SEK 97.0M (2021), SEK 162.5M (2022), SEK 230.9M (2023) and SEK 314.8M (2024), with profit after net financials of SEK 57.4M in 2024 on 44 employees. Those are not estimates. Mullvad is a Swedish aktiebolag, so it files audited accounts with the companies registry every year, and the whole series can be read off them. Doing that turns up three things none of the summaries mention. Revenue grew 7.7x in four years while the price stayed at EUR 5 a month, unchanged since 2009. The company stopped accepting new recurring subscriptions in June 2022, and turnover then grew +94% over the following two years. And the one data aggregator ranking on page one for this question puts Mullvad at "$4.1M per year", which is roughly the company's 2020 figure and about eight times below the 2024 filing.

The number on page one is four years stale

Search mullvad revenue and the first independent-looking source is Growjo, which states that "Mullvad VPN's estimated annual revenue is currently $4.1M per year." ZoomInfo agrees in spirit, listing revenue as "<$5 Million."

Mullvad VPN AB's filed turnover for the year ended December 31, 2024 is SEK 314,810,000. At the cross-rate Mullvad itself publishes (see the currency note below), that is roughly USD 33.2M.

The aggregators are not slightly out. They are out by a factor of about eight. And the shape of the error is specific: USD 4.1M is almost exactly Mullvad's 2020 turnover of SEK 40.7M, or USD 4.3M. Whatever those estimates are tracking, it stopped updating around the time the company was one seventh of its current size.

The rest of page one is not much help either. A Reddit thread quotes "revenue for 2022 was 162 466 000 sek," which is exactly right and exactly one year out of a five-year series. A 2021 Hacker News comment correctly notes that Mullvad "transferred Mullvad to the subsidiary Mullvad VPN AB in 2020" and gives approximate 2020 figures. Wikipedia carries no revenue figure at all.

So the position on August 24, 2026 is that a company legally required to publish audited accounts every year, in a registry anyone can search for free, has no correct current revenue figure anywhere on the open web.

Why every estimate of this company is wrong

There is a clean structural reason, and it is worth understanding because it breaks aggregators on a lot of European companies.

Mullvad's revenue lives in two different legal entities, and it moved between them.

The parent is Amagicom AB, org.nr 556783-9807, registered in 2009 and directly owned by founders Fredrik Strömberg and Daniel Berntsson. The operating company is Mullvad VPN AB, org.nr 559238-4001, registered on January 24, 2020.

The VPN business was moved into the new subsidiary in 2020. You can watch it happen in the parent's own accounts. Amagicom's turnover for its extended financial year ending December 2020 was SEK 25.5M. The following year, 2021, it was SEK 89,000. Not 89 million. Eighty-nine thousand.

Amagicom is now a holding company. Its profits come almost entirely from financial income, meaning dividends received from the subsidiary: SEK 15.2M in 2021, SEK 40.1M in 2022, SEK 17.4M in 2023.

So there are three wrong answers available to anyone scraping this company, and only one right one. Look up "Mullvad" and you may land on the parent, which now trades almost nothing. Look up the subsidiary too early and you catch a company that did not exist before 2020. Only the operating subsidiary's post-2020 accounts describe the actual business.

Mullvad revenue by year, 2020 to 2024

From Mullvad VPN AB's filed annual accounts. All figures in thousands of Swedish kronor, as filed.

Four notes on reading this table honestly.

2020 is a stub year. Mullvad VPN AB was registered on January 24, 2020, so its first accounting period runs from that date to December 31, 2020, about eleven months, and it covers a business being handed over mid-flight. The +138.4% for 2021 is therefore flattered. Every year from 2021 onward is a clean twelve months.

2024 turnover includes SEK 10.9M of other income. Net sales specifically were SEK 303,911,000. Every prior year is essentially all net sales, so the underlying net-sales growth for 2024 is +31.6% rather than +36.3%.

Profit after net financials went flat in 2024 at SEK 57.4M against SEK 57.8M in 2023, even as turnover rose SEK 84M. Operating costs went from SEK 174.6M to SEK 257.2M. Net margin fell from 25.0% to 18.2%. This is a company that spent its 2024 growth rather than banking it.

These are the most recent accounts on file. As of August 2026, the registry shows nothing later than the year ended December 2024, so there is no published 2025 figure yet.

The price has not changed since 2009

Here is the part that should bother anyone who has ever sat in a pricing meeting.

Mullvad's pricing page charges EUR 5 per month. One month, one year, or one decade, all at the same rate. In its own words: "This rate hasn't changed since, well, ever! We're still offering privacy at the same rate as when we launched in 2009."

There is no annual-plan discount, because there is no annual plan, just a longer top-up at the same monthly rate. There is no free tier. There is a 10% discount for cryptocurrency payments, "due to lower fees and less administration," and that is the only discount that exists. On whether they ever run sales, the page says: "Nope. Not for Christmas or any other holiday or event that you can think of."

Now go further. On its trustworthy VPN page, Mullvad confirms it runs no affiliate programme, holds no ownership stake in any VPN review site, and has turned down media offers to buy placement in top-ten rankings. It states plainly that "all revenue comes from paying VPN customers."

And it does not have accounts in the normal sense. Signing up generates a random 16-digit number. No email, no name, no password. You can pay by posting cash in an envelope.

Line up the growth levers a subscription business is normally told to pull, and Mullvad has switched off essentially all of them: no free tier, no trial-to-paid funnel, no annual discount, no price increases, no promotions, no affiliates, no paid placement, no email list, no lifecycle marketing, no upsell tiers, no usage-based expansion. Turnover still went from SEK 40.7M to SEK 314.8M in four years.

They deleted their own MRR and grew 94%

The single most interesting decision in the file is dated June 20, 2022.

On that day Mullvad published a post announcing it would stop accepting new recurring subscriptions: "In order to store less data we will no longer accept NEW PayPal and Credit card subscriptions (recurring payments). One-time payments are not affected."

Read that as an operator. They deliberately destroyed automatic renewal. Every customer now has to remember, unprompted, to come back and top up an anonymous 16-digit number, or the service simply stops. No dunning, no failed-payment recovery, no card-updater, no involuntary-churn playbook, because there are no cards on file to recover. They gave up predictable recurring revenue on purpose, to hold less personal data.

The two full years after that decision:

• 2022 turnover: SEK 162.5M
• 2023 turnover: SEK 230.9M, +42.1%
• 2024 turnover: SEK 314.8M, +36.3%

+93.8% over the two years following the removal of recurring billing.

This is not a claim that killing subscriptions caused the growth. Mullvad also released Mullvad Browser with the Tor Project on April 3, 2023, and got a substantial reputational windfall on April 18, 2023 when Swedish police arrived at its Gothenburg office with a search warrant and left with nothing, because there was nothing stored to take. Both plausibly moved more revenue than the billing change.

The claim is narrower and harder to argue with: removing recurring billing did not stop this company from nearly doubling. For a business whose entire customer promise is "we hold nothing about you," the friction of manual renewal appears to have been priced in by customers as evidence rather than paid as a cost.

Roughly how many people pay for this

Mullvad does not publish subscriber numbers. But a single flat price with no tiers makes the arithmetic unusually clean, which is not true of almost any other subscription company.

Net sales of SEK 303,911,000 in 2024, divided by EUR 5 per month at Mullvad's own published rate of SEK 55.32 to EUR 5, gives about 5.5 million paid subscriber-months, or roughly 458,000 average paying subscribers across 2024.

Treat that as an estimate with a stated direction of error, not a fact:

• It is a floor. The 10% crypto discount means some subscriber-months were bought for less than SEK 55.32, so the true count is somewhat higher.
• It uses one exchange rate across all years. The euro price is fixed but the kronor reporting currency is not, and SEK weakened against EUR over this period, which understates the earlier years more than 2024.
• It assumes essentially all net sales are VPN subscriptions, which the company's own "all revenue comes from paying VPN customers" statement supports.

On the same basis the series runs roughly 61,000 (2020), 146,000 (2021), 245,000 (2022), 348,000 (2023) and 458,000 (2024) average paying subscribers.

Where the money actually goes

Two numbers from the balance sheet are worth an operator's attention.

Payroll. In 2024 Mullvad paid SEK 2,772,000 to its board and CEO and SEK 37,860,000 to everyone else, across 44 employees. That is roughly SEK 923,000 per head in gross salary, and a striking ratio: the entire board and chief executive cost about 7% of the wage bill. Turnover per employee is about SEK 7.2M, or roughly USD 754,000 per person.

Dividends. The proposed dividend was SEK 10M for 2021, SEK 10M for 2022, SEK 30M for 2023 and SEK 30M for 2024. That is SEK 80M distributed out of the operating company across four years, against cumulative profit after net financials of SEK 178M. The founders own the parent directly, so this is a genuinely bootstrapped company paying its owners real money, not accumulating a war chest. Cash at bank at the end of 2024 was SEK 73.4M, down from SEK 89.7M a year earlier, with equity of SEK 50.6M against total assets of SEK 197.8M.

A currency note. Every figure above is stated in Swedish kronor, because that is the currency the accounts are filed and audited in and the only figure that is authoritative. Where I have given a US dollar equivalent I have used a single rate throughout, derived from the cross-rates Mullvad publishes on its own pricing page in August 2026 (EUR 5 = SEK 55.32 = USD 5.83, so about SEK 9.49 to the dollar). That is a comparability device so the years can be read against each other. It is deliberately not each year's average rate, and dollar figures for 2020 to 2023 will not match a historically converted series.

What an operator can actually take from this

Three things, and the last one is the uncomfortable one.

Filed accounts beat estimates, and almost nobody looks. Every number in this piece came from a free public registry search that took a few minutes. Meanwhile the two paid data products ranking on page one for this company's revenue are wrong by a factor of eight. If you are benchmarking against a European competitor, or sizing a market, or quoting a rival's revenue in a pitch deck, the audited filing very likely exists and very likely contradicts the estimate you were about to use. The same discipline works on Ecosia, which publishes its income every single month and is still misquoted by more than half.

Corporate structure is the usual reason a number is wrong. Mullvad is not hiding anything. It is simply a two-entity group where the trading business moved subsidiaries in 2020, and that single fact defeats every scraper pointed at it. If your own company has a holding structure, an operating subsidiary, or a rename in its history, whatever the internet says about your revenue is probably attached to the wrong entity.

Pricing discipline is not the same as leaving money on the table. This is the hard one. Mullvad has held EUR 5 for seventeen years while inflation ate a substantial part of it. Conventional advice says that is negligence. The counter-evidence is a 7.7x in four years at a 29% peak net margin, and the mechanism is legible: the unchanging price is not a pricing decision, it is the product. A privacy company that raised prices opportunistically, ran Black Friday promotions and paid affiliates would be telling customers something about itself that no audit report could undo. What looks like the absence of a pricing strategy is a pricing strategy, and it is the one the customer is buying. That is the same logic Kagi is testing from the other direction by charging for search outright rather than selling attention.

Not every business has a promise that a price rise would break. If yours does, the filing above is what that constraint is worth after four years.

Frequently asked questions