Removing the background from a photo is probably the most repeated edit in the world: for an online store, a WhatsApp sticker, a profile photo or an app icon. Most popular tools solve it by uploading your image to a server, processing it there and sending back the result — which means your photo travels, is stored who-knows-how-long and is usually limited by a paywall. There's another way: process it right in your browser, without the file leaving your device.
A neural network trained to separate "subject" and "background" analyzes
the image and generates a mask automatically. It's the best starting
point for photos with people, pets or products on reasonably
distinguishable backgrounds. In PhotoCut Studio
this is the "AI Cutout" button (shortcut I):
the model downloads once, runs in your browser with WebAssembly and works
even offline after the first visit.
When the AI gets confused — translucent products, very busy backgrounds, several subjects — the classic GrabCut algorithm is still unbeatable: you draw a rectangle around the subject and the algorithm estimates which pixels belong to the foreground by analyzing color and contrast. It's deterministic and predictable: if the result isn't right, you adjust the rectangle and you're done.
No automatic method is perfect at hard edges. That's why the right workflow always ends with two brushes: keep (marks areas the cutout removed by mistake) and remove (erases leftover background). With zoom on the canvas you can work at pixel level.
Ctrl/Cmd+V from the clipboard).
We cut out a real image (1201 × 1500 px) with the AI Cutout and, meanwhile, kept the browser's network tab open. We confirmed what the local method promises: there was no upload of the image to a server — the u2netp model (4.6 MB) downloads once and from then on all the compute happens on your device.
As for the result: the automatic left the silhouette ready in one go, and we only had to run the remove brush over a fringe on the edge where the background had a tone similar to the clothing. The "after" image above is that export, with the background 100% transparent.
| Format | When |
|---|---|
| PNG | The standard with transparency: stickers, logos, icons, print. |
| WebP | Transparency with much smaller files — ideal for the web. |
| JPG | No transparency: only if you apply a color or image background. |
ID documents, photos of minors, unreleased products, client material under NDA… a good part of what gets cut out daily is sensitive information. If the tool processes the image in your browser, that question disappears: the file never leaves your device, there's no server to hack or retention policy to read. That's the underlying technical difference between PhotoCut Studio and cloud cutout services.
PNG or WebP. Both keep the alpha channel; WebP produces much smaller files for the web. JPG does not support transparency, so use it only if you apply a color or image background.
Fall back to the interactive box (GrabCut): draw a rectangle around the subject and the algorithm decides the foreground by color and contrast. Then refine with the keep/remove brushes at zoom.
No. The whole cutout runs in your browser with WebAssembly and a local AI model. The file never leaves your device — there's no server to hack or retention policy to read.
Yes, if you start from a high-resolution image and refine the hair with a small brush and zoom. A moderate edge feather of 1–3 px keeps the outline natural.
Remove an image's background →Why many "free" tools aren't quite free, and how to cut out with no watermark, no account and no usage limit.
Automatic cutout plus brushes to isolate one or several people and keep hair and outlines clean.