How to cut out an object from a photo
Open the photo, press Start selecting, and click the object. The AI outlines it at once. Click to add parts or Alt-click to take parts away, then save a see-through PNG or copy it.
Open a photo
Drop it on the box, choose it, or paste it. JPG, PNG, WebP, GIF, BMP and AVIF open, up to 24 megapixels.
Start selecting
The first time, the AI model (SlimSAM, 40 MB) and its runtime download once. The model then reads the photo, a few seconds, and from there each click is answered in a fraction of a second.
Click the object
One click usually finds the whole thing. If it took too little, click a part it missed; if it took too much, Alt-click the part to leave out.
Save or copy
The cut-out is made at your photo's full size, with the background remover's edges for hair and soft outlines, and cropped to the object if you like.
Adding and taking away
Segment Anything reads the photo once and then turns each set of clicks into an outline. With one click it is unsure whether you mean a shirt, the person or the whole group, so it makes three guesses and shows the one it is surest of. A second click tells it which.
| To | Do this |
|---|---|
| Select | Click the object |
| Add a part | Click it (or pick Add, then click) |
| Leave a part out | Alt-click or right-click it (or pick Take away, then click) |
| Undo a click | Ctrl+Z, or Undo click |
Where does the photo go?
It stays on your device. The AI model downloads from this site once and runs right in your browser, on your graphics card if it can, otherwise on your processor.
- Downloaded once, then kept in your browser for next time
- Saved without metadata like location or camera
Specs and browser support
| Input | JPG, PNG, WebP, GIF (first frame), BMP, AVIF, up to 24 megapixels |
|---|---|
| Output | PNG with transparency, or PNG on a colour; at full size or cropped to the object. Copy puts the PNG on the clipboard |
| Model | SlimSAM-77 (Apache-2.0), a pruned Segment Anything, 40 MB; onnxruntime-web (MIT), 14 or 17 MB |
| Speed | Reading the photo: a few seconds. Each click: about a tenth of a second |
| Browsers | Current Chrome, Edge, Firefox and Safari on a computer. WebGPU in Chrome and Edge |
| Price | Free, no watermark, no account |
Questions
No. The AI model is downloaded to your browser and runs there. Your photo stays on your device.
The background remover keeps the photo's main subject on its own. Here you choose what to keep by clicking: one item among many, part of a scene, or something the background remover would not count as the subject.
Yes. Click each one; every click adds to the selection. If it joins them with what lies between, Alt-click the gap.