Resize an image to the exact pixels you need

You can make it bigger. You cannot make it sharper. A 400 × 300 photo holds 120,000 pixels of real detail, and asking for 1600 × 1200 does not go back to the camera for the missing ones — the browser invents them by averaging the neighbours it already has. The result is a larger, softer version of the same picture, and edges that were crisp turn slightly blurred. This tool will still do it, because sometimes a system demands a minimum size and a soft image that uploads beats a sharp one that is rejected. It just says plainly what is happening rather than pretending the detail came back. If you need genuine detail, the only fix is a bigger original.

Dimensions are how many pixels the image is made of — 1920 × 1080. File size is how many bytes it takes to store them once compressed — 480 KB. They move together but they are not the same thing, and the confusion causes real problems. A 4000 × 3000 screenshot of a plain white document can be a smaller file than a 1000 × 750 photo of a forest, because flat areas compress to almost nothing and leaves do not. So when an upload form rejects your file, read which one it is complaining about: "maximum 2 MB" is a file size limit, and "maximum 1600px" is a dimension limit. Resizing fixes both, but if the dimensions are already fine, compressing is the smaller change.

Resizing keeps the whole picture and changes how many pixels describe it — everything in the frame stays in the frame, just smaller or larger. Cropping keeps the pixel density and throws away the parts of the frame you do not want. If a form wants a square profile picture and your photo is a wide landscape, resizing alone cannot help: squeezing a 3:2 photo into a 1:1 box either distorts faces or leaves bars. That is a cropping job. If the shape is already right and the file is simply too big, that is a resizing job.

It keeps the width and height in the same proportion as the original, so typing one dimension fills in the other. With it unlocked you can type any two numbers you like, and the image is stretched to fit them — a 4:3 photo forced into a 1:1 box makes everyone in it look narrower. There are legitimate reasons to unlock it, mostly involving textures and background tiles, but for anything with a face or a straight line in it, leave it on.

Making an image smaller discards pixels, which is unavoidable and usually invisible at the size you end up viewing it. The bigger issue is the format. JPEG and WebP are lossy, so every save throws away a little more detail — resave the same JPEG ten times and you can see the damage around edges and in flat colour. PNG is lossless, so a PNG re-saved a hundred times is identical to the first. If you are going to make several passes at an image, do the intermediate steps as PNG and only encode to JPEG once, at the end.

PNG for logos, screenshots, line art and anything with transparency, because it is lossless and keeps hard edges clean. JPEG for photographs, and for anything old or fussy that has to open the file. WebP when you want the smallest result and the destination is a modern browser — it beats both on photographs and supports transparency too. One thing to know: the quality slider disappears when PNG is selected. That is not an oversight. PNG has no lossy mode at all, and the browser ignores the quality value entirely, so showing the control would be pretending it does something.

No, to both. There is no server in this. Your browser reads the file, redraws it on a canvas and encodes a new one on your own device, which is why the tool also works with your internet disconnected once the page has loaded. Nothing is written to localStorage either — not the image, not your settings — so closing the tab leaves nothing behind. The one third-party request the page makes is for the advertising, and that happens whether or not you open a file.

Yes, as a side effect of how it works rather than as a feature bolted on. The image is redrawn as raw pixels before being encoded again, so camera model, timestamps, editing history and GPS coordinates do not survive into the file you download. That is usually exactly what you want before posting a photo publicly — a holiday picture straight from a phone often carries the coordinates of where it was taken. If you actually need the metadata kept, keep the original file too; there is no way to preserve it through a re-encode.