Why Free Online Photo Editors Show Ads: RetouchPi Measures It

A browser editor can stay alive without charging, provided something else pays. Display advertising is the oldest and most direct answer: ad networks pay per impression and per click, and the product rents out part of its layout. For the reader that model has one direct consequence - part of the screen you are retouching on belongs neither to you nor to the editor, but to whoever is bidding highest right now.

• Servers and bandwidth are a fixed monthly bill: the editor runs on your machine, but the site, the accounts, the assets and the updates need hosting.
• Decoders and development are a one-off cost amortised for years: readers for shells like PSD, TIFF and RAW take months to write once.
• The ads themselves carry support costs: one product wrote four self-help documents about ads alone (ads not showing, being accused of blocking when you are not, malicious ads, how ads are displayed). That count is itself evidence of the search volume behind them.

One of the busiest free online editors states it plainly in its own support docs: ads may cover between 5% and 15% of the screen; a regular unit is 300x250 px, placed on the right side of the screen, multiple times; when an ad blocker is detected, various JPG images are placed instead of the ads, some asking you to buy its premium tier, some leading to small games. A second article from the same product admits malicious ads have appeared: fullscreen ones, ones that play sound, ones that open a new website without a click - and even a single report can mean that unit was already shown to tens or even hundreds of users.

The method is not exotic: on 2026-10-10, in the bundled Chromium at a 1280 x 800 viewport, a real photo (900 x 676) opened for real, and once the canvas had actually painted and the tab label had actually changed to the new file, nine ad-slot selectors were counted in the DOM. Hits: zero. Then the editor's single static page was grepped at the source: adsbygoogle and googlesyndication each appear zero times.

The most honest paragraph in that product's docs is the blocker one: when a blocker is detected the slot does not stay empty, it is filled with JPGs - some asking you to buy the premium tier, some leading to small games. What the blocker buys, in other words, is not a clean layout but a different content. The reader believes they are blocking ads while actually playing a round of swapped cards with the product, and what comes up next is not known in advance.

Our version has no such round: the editing surface carries no ad slot to begin with, so a blocker finds nothing to block inside the editor and there is no downgrade to fall into. That sentence holds only on the editing surface, though - other pages of the site, including the one you are reading, do carry ad slots, and on those pages a blocker blocks real ads, and blocked means blocked.

• The malicious class: fullscreen, playing sound, jumping without a click. One product's support docs admit that behind a single report there can already be tens or hundreds of ruined sessions.
• Layout shift: ad slots arrive at unpredictable times and push the canvas and panels around, so the button you are aiming at can move before your click lands.
• Network: requests to ad domains compete with your editing requests over the same connection, so what slows down is not the ad itself but the few hundred milliseconds you wait for the canvas to answer.
• Privacy: what an ad network learns is which page you were on, how long you stayed and what you clicked - a record that says nothing about your photo and a lot about your habits.

None of these costs are written into the word "free". They are written into the three minutes you spend on an ID photo, a little at a time, with no invoice.

• Open the editor and open a photo of your own so the canvas really paints - counting ad slots on an empty canvas proves nothing, because ads often mount only after content appears.
• Open the console with F12 and run document.querySelectorAll for each of the nine selectors in the table above, counting the hits.
• Filter the network panel by the googlesyndication and doubleclick domains and edit for three minutes, watching whether either domain is ever requested.
• Search the page source for adsbygoogle: an ad script has to be named in the source to load, so a source without it is a runtime without it.

With all four done, "ad-free" stops being a slogan and becomes a reading you took yourself. It is the same set of moves we ran on our own version, plus one full-screen capture kept as evidence.

• No site-wide ad-free claim: the landing page's right rail carries ad slots and the whole column disappears only under the ad-free tier; this very page lives on the site and has its own slots.
• No claim that ads are evil: they are how the free ledger is kept. All we do is open the ledger and let the reader decide whether a clean layout is worth paying for once.
• No claim about other products' numbers: 5%-15% and 300x250 are that one product's own stated figures; we quote them, we do not audit them, and we do not generalise them to every free editor.
• No offline claim: the editor is a web page and there is no editor without a network. Ordinary retouching stays on your machine; AI generation and AI edits are the one step that touches a server.