OpenStreetMap Foundation revenue hit a record GBP 470,211 in 2024, and so did its loss

Quick answer (September 2026). The OpenStreetMap Foundation's turnover was GBP 470,211 in 2024, against GBP 445,794 in 2023, GBP 374,424 in 2022, GBP 354,561 in 2021 and GBP 226,273 in 2020. The published series runs back to GBP 88,681 in 2013, so 2024 income is 5.3 times its 2013 level, a compound annual growth rate of 16.4% over eleven years. 2024 was the Foundation's highest-earning year on record and its worst loss on record: a deficit of GBP 91,775, which cut reserves from GBP 700,781 to GBP 609,006. Every figure here comes from the Foundation's own filed accounts. No 2025 accounts had been published as of 7 September 2026.

Now the part that is not anywhere else.

The OpenStreetMap Foundation has filed a set of accounts every year since 2014, and its wiki carries income and balance sheets going back to 2006. That is eleven statutory statements, seven budgets, eleven treasurer's reports and a decade of board minutes. Nobody appears to have put the revenue line from all of them into one table.

I did, and then I checked the arithmetic. All twenty-two turnover columns across the eleven statements foot exactly, to the pound. Every year that is printed twice, once as the current year and again as the prior-year comparative in the next statement, agrees with itself: ten out of ten, at the total and across sixty-nine individual line-item cells. The Foundation's accounts are internally consistent to the pound across a decade, and that deserves saying before anything else.

What the exercise did surface is a record year that is also a crisis year, a source of income that has quietly grown seventy-seven fold, a single anonymous Bitcoin donation that was nearly half of one year's income, and a set of accounts that is published but is missing from the page that lists the accounts.

OpenStreetMap Foundation turnover, 2013 to 2024

The OpenStreetMap Foundation is a company limited by guarantee registered in England and Wales, number 05912761, with a registered office at St John's Innovation Centre, Cowley Road, Cambridge. It is not a charity, and its accounts are unaudited: it claims the small-companies audit exemption under section 477 of the Companies Act 2006 every year. All figures below are in pounds sterling, taken from the Foundation's own published financial statements.

The surplus and reserves columns start at 2017 because that is as far back as the balance sheet figures could be read reliably from the filed PDFs. Turnover goes back to 2013, which appears only as the comparative column in the 2014 statement.

Two things stand out. The first is how lumpy this is. A project that most people assume ticks along on small donations has swung by more than 100% in a single year twice, in 2016 and 2018, and has fallen by a third once, in 2019. The second is that the surplus column and the reserves column reconcile perfectly: every year's reported surplus, added to the previous year's closing reserves, produces the next year's closing reserves exactly, seven times out of seven. That is a basic check, and it passes.

2024: the best year and the worst year at the same time

2024 is the Foundation's highest-earning year in the published record, at GBP 470,211. It is also the worst loss in the published record, a deficit of GBP 91,775. Turnover rose by GBP 24,417 while cost of sales rose by GBP 98,133, from GBP 76,485 to GBP 174,618, and administrative expenses rose by a further GBP 25,114.

The Foundation's own treasurer put it plainly in the report to the 2025 annual general meeting.

The Foundation's treasurer, writing to members in the 2025 report: "The Foundation has made a big loss this year and this is, and should remain, be alarming."

He goes on to say that the preliminary balance for the following year again looks like a surplus of around GBP 38,000, and that this is "not real, as in particular income and to lesser extent expenses are distributed unevenly over the year", and that the board now forecasts "again a big loss for this year". His three stated movements between 2023 and 2024, turnover up about GBP 25,000, cost of sales up almost GBP 100,000 and administrative expenses up about GBP 25,000, all check out against the filed statement.

Cash at bank fell from GBP 640,940 to GBP 574,537 over the same year. Reserves at the end of 2024 stood at GBP 609,006, which is roughly thirteen months of 2024's total cost base of GBP 562,038. This is not an organisation in immediate danger. It is an organisation whose costs grew faster than a record income year could cover, and whose treasurer said so in public.

Where the money actually comes from

The composition has changed more than the total has. In 2014 corporate membership brought in GBP 3,000, about 2.5% of income. In 2024 it brought in GBP 232,249, 49.4% of income. That is a seventy-seven fold increase, and it is the real structural story of OpenStreetMap Foundation finance over the last decade.

Individual membership, the thing most people picture when they picture funding a mapping project, is GBP 20,130, or 4.3%. Merchandising is sixty-nine pounds. The Foundation that runs the map is now, financially, a corporate membership body with a conference attached.

The year a stranger paid for nearly half of it

The 2018 accounts carry a line that appears in no other year: "Donations - Pineapple fund", GBP 186,790. That is 46.5% of the whole of 2018's turnover from one source.

The Pineapple Fund was an anonymous Bitcoin philanthropy run in 2017 and 2018 by a person known only as Pine, who the Foundation described as one of the 250 largest Bitcoin holders and who pledged to give away 86 million dollars' worth of Bitcoin. The Foundation announced the donation on 11 January 2018, describing it as 18 Bitcoin, "currently worth over EUR 200,000".

The sterling figure in the accounts and the euro figure in the blog post are not directly comparable, because what a Bitcoin gift is worth in the books depends on when it was converted, and I have not tried to reconcile them. Both are the Foundation's own published numbers and I am reporting them as published.

There is a second detail in that announcement worth noting. Two weeks before the Pineapple gift, the Foundation had received a separate anonymous donation of "over 2.3 BTCs, worth EUR 30,000 (GBP 27,000)", from someone else entirely. The 2017 accounts show total donations of GBP 28,668 for the whole year. The timing and the size are consistent with that single gift being almost the entirety of 2017's donation income, though the accounts do not break it out and I am flagging that as an inference rather than a fact.

One caveat on sourcing: pineapplefund.org, which the Foundation's own blog post links to for the grantee list, no longer belongs to the Pineapple Fund. The domain has lapsed and now serves unrelated commercial content in Korean. The fund's public record of who it gave money to is gone, so the donation cannot be verified at the giver's end. The Foundation's own contemporaneous announcement and its own filed accounts are what remain.

Why nobody has this number

There are three reasons this series has not been assembled before, and they are all boring and all real.

The most recent accounts are not on the accounts page. The Foundation's Finances wiki page lists "Official Financial Statement" entries for 2023, 2022, 2021 and back to 2015. There is no 2024 entry. The 2024 accounts do exist and are published, but the only link to them sits inside the 2025 treasurer's report, one page away. Anyone who goes to the page that lists the accounts, which is the obvious place to go, will conclude the record stops at 2023.

The statement changed its name twice. The same schedule of line-item income is called the "Detailed Profit and Loss Account" in 2014 and 2015, the "Trading and Profit and Loss Account" in 2016 and 2017, and the "Detailed Income and Expenditure Account" from 2018 onward. Search for one name and two thirds of the record is invisible.

The columns swap sides in 2018. In the statements up to 2017 the prior year is printed in the left money column and the current year on the right. From 2018 the current year is on the left. Read the series without noticing, and every figure before 2018 lands against the wrong year. This one is worth spelling out because it is the trap that makes a plausible, confident, entirely wrong table: I built one, then caught it because 2015 has to be the same number in both the 2015 and 2016 statements, and it was not until the columns were read by position rather than by order. After the fix, the whole chain agrees.

The Foundation is also filed at Companies House, where the accounts are scans rather than text, which defeats copy and paste. Its 2024 filing does not include the income statement at all: it says so on the face of the balance sheet, "In accordance with Section 444 of the Companies Act 2006, the Income Statement has not been delivered". The public register holds the balance sheet and nothing else, which is entirely lawful for a small company and does mean the register alone cannot tell you what the Foundation earned. Cross-checking the Companies House filing against the self-published PDF is worth doing anyway, and it passes: all six balance sheet lines in the 2023 comparative column match the Foundation's own document to the pound.

What the top search result actually says

Search for OpenStreetMap revenue today and the first result is a 2023 Hacker News comment. It says the 2021 financial report showed "turnover of GBP 350,000 with surplus of GBP 160,000 and GBP 740,000 in reserves".

To be fair to it: that comment is right. The filed 2021 figures are turnover GBP 354,561, surplus GBP 159,428 and reserves GBP 742,516. Rounded to the nearest ten thousand, all three of its numbers are correct.

The problem is not accuracy. The problem is that the best answer the open web currently offers to a question about a well-known organisation's finances is one rounded year, buried in a forum thread, now five years out of date, and describing a year that turns out to be the Foundation's best surplus rather than a typical one. Every year since has been published, and three of them have been worse.

Which OpenStreetMap are you actually asking about

A warning that will save you an hour, because the search results mix these up freely. Several separate legal entities carry the OpenStreetMap name and none of their accounts are interchangeable.

• OpenStreetMap Foundation, a UK company limited by guarantee, number 05912761. This is the body that runs the servers and holds the database rights, and it is the only entity in this article.
• Humanitarian OpenStreetMap Team, a separate United States non-profit with its own audited accounts, which is considerably larger in revenue terms and files a US Form 990.
• OpenStreetMap US and OpenStreetMap Ireland, national chapters with their own separate and much smaller accounts. OpenStreetMap Ireland's 2020 revenue, for instance, was EUR 219.30.

If a figure for "OpenStreetMap revenue" seems surprisingly large, check whether it is actually the Humanitarian OpenStreetMap Team.

Method

Seventeen PDFs were downloaded from the Foundation's wiki: eleven statutory financial statements covering 2014 to 2024, five budgets, and the 2017 independent volunteer auditor's report. All seventeen returned HTTP 200, all were text-extractable, and all had distinct checksums. Two further sets of accounts, for 2024 and 2023, were taken from Companies House; those are image scans and were read by rendering each page. Two more budgets, for 2023 and 2026, exist as wiki pages rather than PDFs and were not part of the arithmetic checks.

Line items were extracted by position on the page rather than by reading order, because reading order is unreliable in the pre-2018 layout and puts the two money columns in the wrong sequence. Each year's figures were then checked three ways: the line items were summed and compared against the printed subtotal, each year was compared against its own restatement as a comparative in the following year's document, and the reserves movement was compared against the reported surplus. The counts are twenty-two turnover columns, ten restatement pairs, sixty-nine line-item cells and seven reserves reconciliations, with zero disagreements anywhere.

The one difference found across all of it is a label: "Merchandising and commissions" became "Merchandising and commission" between the 2019 and 2020 statements. The value, forty-two pounds, is identical in both.