QGIS revenue was EUR 428,260 in 2024, and eleven years of its accounts add up to the cent

Quick answer (September 2026). QGIS.ORG's total income was EUR 428,260.48 in 2024, against EUR 370,906.17 in 2023, EUR 214,171.57 in 2022, EUR 233,076.87 in 2021 and EUR 179,221.65 in 2020. The published series runs back to EUR 68,753.32 in 2015, so income is 6.2 times its 2015 level, a compound annual growth rate of 22.5% over nine years. In 2024 the organisation spent EUR 423,473.45 and booked a profit of EUR 4,787.03. Its own approved budget for 2026 plans EUR 552,000 of regular income. Every figure here comes from QGIS.ORG's own financial reports. No 2025 financial report had been published as of 6 September 2026.

Now the part that is not anywhere else.

QGIS.ORG has published a financial report every year since 2014 and a budget every year since 2015, plus an independent auditing report every year since 2017 and a project annual report every year since 2016. That is forty documents. Nobody appears to have added them up.

I did. All twenty-two income and expense columns across the eleven financial reports foot exactly, to the cent. EUR 2,080,331.90 of income and EUR 2,111,554.86 of expenses, hand-assembled tables, eleven years, zero arithmetic errors. Eleven years of hand-assembled tables without a single column that fails to add up, and it deserves saying first.

What the exercise did surface is one year where the paperwork slipped and never got tidied, three small discrepancies in the surrounding paper trail, and a headline number in the 2026 budget that is 12.7% smaller than it looks.

QGIS.ORG income and expenses, 2014 to 2024

QGIS.ORG is a Swiss association (a Verein) registered at Via Geinas 2, CH-7031 Laax. It publishes at qgis.org/community/foundation/finance/. All figures are in euro, taken from the annual financial reports.

Two caveats before anyone quotes this. 2014 is not a full year. That report's own opening line says the period "starts on July 8, 2014, when QGIS.org obtained its own bank account", before which the Swiss QGIS user group handled the money on separate books. So 2014 covers roughly six months and no year-on-year change can be computed from it.

And the biggest income line was renamed. Through 2018 it is "Sponsoring"; from 2019 it is "Sustaining Membership Payments". These are the same line, and QGIS.ORG confirms it in its own budget documents: the 2020 budget labels the row "Sustaining memberships (former sponsorships)". Without that confirmation the series would look like sponsorship collapsing to zero and a new revenue stream appearing from nowhere. It is a rename, not a change of business.

Where the money actually comes from

Corporate and institutional memberships are the engine and have been since 2016. But the last two growth years were driven by completely different things, and the difference matters if you are forecasting.

2023 was a membership year. Income rose EUR 156,734.60, of which EUR 131,765.67, or 84%, was memberships.

2024 was a donation year, and it was one channel. Income rose EUR 57,354.31. Stripe donations rose EUR 71,752.11, from EUR 25,881.98 to EUR 97,634.09. That single line is 125% of the entire increase, because everything else net-declined: memberships added only EUR 3,121.04, while PayPal fell EUR 5,114.68, bank transfers fell EUR 3,428.65, GitHub Sponsors fell EUR 904.33, certificate income fell EUR 1,240.00 and the "Other" line of EUR 7,073.24 vanished entirely. The deltas reconcile to the cent against the printed totals.

The 2024 report does not explain the Stripe jump. The 2026 budget quietly folds all card channels into one line labelled "Credit card donations (payrexx/stripe/Paypal)", which suggests a processor change rather than a fourfold surge in generosity, but QGIS.ORG has not said so and I am not going to assert it.

The number QGIS.ORG has never printed

Every report since 2017 contains the same sentence: approved trainers "can purchase official QGIS.ORG certificates (EUR 20.- per certificate)". The reports print the income from that scheme. They never print how many certificates it represents.

Divide one by the other and all eight years come out as exact whole numbers, which is itself the proof that the EUR 20 price held throughout:

6,095 official QGIS training certificates have been sold since 2017. If a single year had produced a fraction I would not publish this, because the price would have moved. None did.

2020: the year the paperwork never closed

Only seven of the eleven reports print a profit or loss line at all; the 2014 to 2017 reports simply stop after the two totals. Six of those seven reconcile from their own printed figures, in the sense that total income minus total expenses equals the number printed at the bottom. The 2020 report does not.

It prints total income of EUR 179,221.65 and total expenses of EUR 211,611.31. That implies a loss of EUR 32,389.66. The report prints "Loss in 2020: 10,939.90".

The gap is exactly EUR 21,449.76, and the reconciling item is on the page. Immediately below the income total, outside it, sits a line reading "Reserves for tax payments (Reverse charge VAT) for years 2016 to 2019: 21'449.76". That is a release of reserves the 2019 report had created, and the 2019 report's three matching lines (EUR 3,744.00, EUR 7,811.65 and EUR 9,894.11) add to precisely that figure. So the money is real, the bottom line is right, and the presentation is what fails: the published "Total Income 2020" understates the resources actually available that year by EUR 21,449.76.

Two other things about that report have never been fixed. Its status line still reads "approved by financial auditors, but not yet approved by AGM". And its signature block still reads:

Verbatim from the published PDF: "Approved by the AGM 2021 on xxx xx, 2021"

An unfilled template placeholder, live on the QGIS.ORG website as of 6 September 2026. It is the only one of the eleven reports containing the string. Eight reports carry an AGM approval line at all, the 2017 report being the first, and the other seven all name a real date, running from "AGM 2018 on 28 March 2018" through to "AGM 2025 on October 3, 2025".

Three discrepancies in the paper trail

Two auditors, Andreas Voigt of the German QGIS user group and Andreas Vonlaufen of the Swiss one, have signed an independent report every year since 2017. Each report quotes four hard figures: actual expenses, budgeted expenses, actual income, budgeted income. I checked all thirty-two against the underlying documents. Thirty match. Two do not. A third discrepancy sits not in an audit report but inside a budget.

One. The 2020 auditing report opens "With EUR 211,661.31 the total expenses...". The financial report says EUR 211,611.31. Fifty euro apart, one digit. The financial report's expense column adds up to its own printed total across all twenty-one line items, so the report is right and the audit carries the slip.

Two. The 2017 auditing report says income exceeded "the budgeted total income of EUR 105,350.00". The published 2017 budget prints EUR 104,850.00, and its income column adds up to exactly that. The two documents agree on budgeted expenses of EUR 109,850.00 and differ by EUR 500 on income. The audit names its source as "the budget approved by AGM 2017 as of May 8, 2017", which may not be the version now published, so I am not going to say which is authoritative. I am only going to say they differ.

Three, and this one is not the auditors' at all. It is internal to a single page. That same 2017 budget carries a closing remark: "there is a loss of EUR 9'000 budgeted". Its own two columns both foot exactly, to EUR 104,850.00 and EUR 109,850.00, which is a budgeted loss of EUR 5,000. Between one document and one sentence on it there are three different figures for the same quantity: EUR 4,500 on the auditors' income number, EUR 5,000 on the published table, EUR 9,000 in the prose.

None of these changes anything material. All three are checkable in about a minute, which is the point.

Budget against actual, ten years

This is the part I did not expect. QGIS.ORG's forecasting is poor and its bottom line is nearly perfect.

In 2023 income came in 76.6% above budget and expenses came in 78.4% above budget. They spent the surprise, in the same year, almost exactly.

The result is that for four consecutive years the organisation has landed within EUR 5,500 of break-even: +5,412.38, -4,115.84, -4,599.33, +4,787.03. Across those same four years annual income ranged from EUR 214,172 to EUR 428,260, a factor of two between the smallest and largest. In 2024 the net result was 1.12% of income. Whatever the budget says in January, the spending is tuned to whatever arrives.

The EUR 45,000 that was not spent, and the EUR 80,000 that is not income

Two things will trip anyone reading the newest documents quickly.

2024 expenses include EUR 45,000 that went nowhere. Two lines near the bottom read "Creation of Reserves for Website Work: 25'000.00" and "Creation of Reserves for Documentation Work: 20'000.00". That is money set aside for 2025, booked as a 2024 cost. QGIS.ORG's own auditors make the same point, noting that without those two reserves the overshoot against budget would have been "only around EUR 4,000". So actual project spending in 2024 was EUR 378,473.45, not EUR 423,473.45.

And the 2026 budget's headline is not what it appears. The approved 2026 budget prints "Total Income EUR 632,000.00". Above it, it prints "Regular Income EUR 552,000.00". The EUR 80,000 difference is a line reading "Release of reserves from 2024 and 2025".

To QGIS.ORG's considerable credit, that subtotal is printed right there on the page, which is exactly the disclosure the 2020 report lacked. But EUR 632,000 is not a revenue forecast. The 2026 revenue forecast is EUR 552,000, which is 28.9% above 2024 actual rather than 47.6% above it. Quote the wrong one and you overstate the plan by 14.5%.

For completeness: the 2025 budget planned EUR 473,000 of income against EUR 505,850 of spending, a deliberate EUR 32,850 drawdown. No 2025 financial report exists yet, so nobody knows how that landed. On the recent pattern the 2024 report was signed off by the auditors in June 2025 and approved by the AGM that October, so a 2025 report is not late.

What checks out

Because it would be easy to read the above as a catalogue of faults, here is what I tested and could not break.

The 2024 report states that its sustaining members number "132 in total, originating from 31 countries from all continents except Africa and Antarctica". I counted the list. There are exactly 132 entries, split 2 flagship, 8 large, 27 medium and 95 small, which is exactly what the four tier headings claim. They resolve to exactly 31 distinct countries, spanning Europe, Asia, North America, South America and Oceania. Not one African member. Three checkable claims, three correct.

There is a quiet irony in that last one. QGIS.ORG's two employees, a documentation writer and a full-stack developer, are subcontracted through Kartoza, which is South African. The continent with no sustaining members is the one doing a chunk of the work.

The 2024 expense side also names where the money goes with unusual specificity: EUR 46,455.08 and EUR 35,995.54 and EUR 40,394.00 on bug fixing for releases 3.36, 3.38 and 3.40 LT, EUR 42,025.00 on packaging for Windows, Ubuntu and Apple, EUR 37,144.30 on the grants programme across six named grantees, EUR 21,000.00 on upstream Qt improvements, EUR 6,000.00 on an OpenStreetMap Foundation corporate membership, and EUR 4,749.15 on Hetzner and macOS build servers.

And the auditors are visibly doing the work. Their 2024 report flags that Kartoza "charged US$677 as EUR 677" and recommends not correcting it because "the shortfall is small". At the European Central Bank reference rate for September 2024, when the item was posted, EUR 1 bought USD 1.1106, so US$677 was about EUR 609.58 and the currency slip is roughly EUR 67. They found a sixty-seven euro discrepancy in a four hundred thousand euro year, wrote it down, and explained why they were leaving it.

One documentation defect worth reporting

QGIS.ORG's annual reports page says of the project reports: "Note that the report will contain some financial information which is also available as a separate report in the finances section of this web site."

That was true of the first two. The 2016 annual report contains six euro figures and eight references to income; the 2017 one contains two and five. Every annual report since, all seven of them, contains zero euro symbols, zero occurrences of "income" and zero occurrences of "budget". The financial section was dropped after 2017 and the reports shrank to roughly a third of their former length. The sentence describing them was never updated.

If you want QGIS money numbers, the annual reports have not carried any for seven years. Go to the finance page.

What to quote

For QGIS revenue, quote EUR 428,260.48 for 2024 and EUR 370,906.17 for 2023, and say "income" rather than revenue, since roughly 62% of it is membership fees and most of the rest is donations.

For the 2026 plan, quote EUR 552,000, not EUR 632,000.

For 2024 spending, quote EUR 378,473.45 if you mean money spent on the project, or EUR 423,473.45 if you mean the accounting total including the EUR 45,000 moved to reserves.

For 2020, quote the loss of EUR 10,939.90 rather than the EUR 32,389.66 the printed totals imply, and know why the two differ.

And if you are comparing this to other open-source foundations, the useful frame is the one the membership list itself supplies. QGIS.ORG maintains a desktop GIS whose paying supporters include national agencies, regional and municipal governments, universities and consultancies across 31 countries, and it does that on an income smaller than a mid-sized agency's payroll, with two subcontracted employees and a treasurer who has hand-built and balanced eleven consecutive sets of books.