Crop an Image Online
Drag a selection over your image and export just that region. Aspect ratio presets cover the shapes platforms actually demand — square for a profile picture, 9:16 for a story, 16:9 for a video thumbnail — and the crop is measured in the image's own pixels, so what you export is exact even though the preview on screen is scaled to fit.
Use this without the search next time. Prathom Workbench puts Prathom's tools in your toolbar.
Add to Chrome — freeWhat it does
- Drag to move, handles to resize
- Locked aspect ratio presets
- Live pixel dimensions and ratio readout
- PNG, JPEG, or WebP output
- Nothing uploaded at any point
How to use Image Cropper
- 1
Open an image
Drop a file in or click to browse. The whole image starts selected, so you can drag inwards from there.
- 2
Choose a ratio, if you need one
Pick a preset to lock the shape. Selecting one recenters the selection at the largest size that fits, so it is always a valid starting point.
- 3
Position the selection
Drag inside the box to move it. Drag any of the eight handles to resize. With a ratio locked, the other dimension follows automatically.
- 4
Crop and save
Press Crop to render the region, check the result, then save it. Choose PNG to keep transparency or JPEG for a smaller photograph.
How it works
The selection is stored as four numbers in the image's own coordinate system: left, top, width, height. The preview is drawn at whatever scale fits your screen, and every pointer movement is divided by that scale before it changes the rectangle. Keeping the crop in image pixels rather than screen pixels is what makes the exported region exact — a preview scaled to 18 percent of a large photo would otherwise quietly turn a full-resolution crop into a small one.
Handle dragging works by converting the rectangle into two opposite edges, applying the movement to whichever edges the handle touches, and rebuilding. That sounds like a detail, but it is what makes dragging the right handle leftwards past the left edge behave sensibly: the edges swap and you get a valid rectangle, rather than a negative width and an invisible selection.
With a ratio locked, one axis leads and the other is derived. Corner and horizontal handles drive from the width; top and bottom handles drive from the height. The rectangle is then re-anchored to the edge you were not dragging, so the opposite corner stays pinned instead of the whole box drifting across the image as you resize.
The export itself draws the selected source region into a canvas of exactly that size and encodes it. There is no scaling step, so the pixels you selected are the pixels you get.
When you'd use this
Profile pictures and avatars are the most frequent case, because almost every platform wants a square and almost no camera produces one. Locking the square preset and dragging up to center a face takes seconds.
Social formats are the second. Stories and reels want 9:16, feed posts often want 4:5, video thumbnails want 16:9. Cropping to the right shape yourself gives you control over what gets cut; letting the platform crop automatically usually means the top of someone's head disappears.
Removing surroundings is the third and least glamorous: trimming a screenshot down to the relevant panel, cutting the edge of a desk out of a product photo, or straightening the framing of a document photographed at arm's length. All of it is faster than opening an image editor for one operation.
Cropping versus resizing
These are different operations that are easy to confuse, and the difference matters when a platform gives you a required size.
Cropping changes which part of the image you keep. The remaining pixels are untouched, and the output dimensions are however large that region happens to be.
Resizing changes how many pixels the image has. Everything stays in frame and each pixel is recalculated from its neighbors.
A requirement like "1080 by 1080" usually needs both: crop to a square so nothing is distorted, then resize that square to 1080. Doing only the second would squash a rectangular photo into a square, which is the error that produces the stretched-looking profile pictures you see everywhere.
Examples
Making a square profile picture from a portrait photo
Choosing the square preset immediately produced a 3024x3024 selection — the largest square this portrait can contain — positioned centrally, which on a portrait photograph usually lands on the torso rather than the face. Dragging it upward before cropping is almost always the next step.
Pulling a 16:9 thumbnail out of a screenshot
Because the ratio was locked, dragging the corner handle horizontally set the height automatically and the readout stayed at 16:9 throughout. Without the lock this would have needed arithmetic, and any error would only become visible after upload as black bars.
Frequently asked questions
Does cropping reduce the image quality?
Not the part you keep. Cropping copies the selected pixels unchanged into a new image, with no resampling or scaling involved, so the retained region is pixel-identical to the original. The only quality consideration is the output format — saving as JPEG applies its usual compression, whereas PNG keeps everything exactly.
Why does the preview look smaller than my actual image?
Large photographs are scaled down to fit on screen, otherwise a 4000-pixel image would not be visible at once. The crop rectangle is stored in the image's real pixel coordinates rather than screen coordinates, and every drag is converted before it is applied, so a selection that looks like 500 pixels on screen can correctly export as 3024.
Can I crop to an exact pixel size like 1080 by 1080?
Indirectly. Lock the square ratio and drag until the dimensions readout shows the size you want — it updates live as you drag. If you need a guaranteed exact figure, crop to the right shape here and then use the image resizer with Exact size, which will scale the cropped region to precisely those dimensions.
What happens to transparency when I crop a PNG?
It is preserved if you export as PNG or WebP. Choosing JPEG as the output fills every transparent pixel with white before encoding, because JPEG has no transparency channel and unfilled pixels would otherwise come out black. For logos and cutouts, keep the output as PNG.
Is there an undo?
Not as a history stack, but nothing is destructive: the original decoded image stays in memory untouched and every crop renders from it fresh. Changing the selection and pressing Crop again replaces the result. Only the file you save to disk is permanent, and your original file is never modified.
Further reading
- Cropping a JPEG Re-Encodes the Part You KeptCropping feels subtractive — you are taking pixels away, not changing them — so the part you keep should arrive untouched. In a JPEG it usually does not, and the reason is the same one that makes rotation lossy.
- GIF Transparency Has Exactly One Bit, and It Shows on Every EdgeA PNG can record a pixel as 37% transparent. A GIF cannot — each pixel is either fully visible or fully invisible, and there is nothing in between. On a shape with a soft anti-aliased edge, that difference is measurable, and it is why a logo saved as GIF gets a jagged fringe that no amount of care in the original file prevents.
- Six Ways to Shrink an Image, and What Each One CostsEvery image editor has a dropdown of resampling filters with names like Bicubic, Lanczos and Nearest Neighbor, and almost nobody changes it. The differences between them are real, they are measurable, and which one is right depends on what is in the picture rather than on which sounds most advanced.