EveryToolAI
Index

Image Resizer

This resizer produces an image at the exact pixel dimensions you ask for. Enter a width, a height, or both, choose whether to fit the whole picture inside those bounds or fill them and crop, and the new image is drawn in your browser and downloaded. Nothing is uploaded and the original file is never modified.

Resizing is not the same job as compressing

These two get conflated constantly, and the difference decides which tool you want. Compressing answers "make this file smaller than 2 MB" — the dimensions are a means to that end. Resizing answers "make this exactly 1080 by 1080" — the number came from somewhere else and is not negotiable.

The second question is usually asked by a form. A marketplace wants a square thumbnail. A conference wants a headshot at a fixed size. A CMS crops anything that is not 16:9. A print shop rejects anything under a dimension. In all of those the target is dictated, and a tool that gets you close is no use.

So this tool takes dimensions literally and tells you exactly what you are going to get before you press anything. If your goal is really a smaller file rather than a specific size, the image compressor linked below is the better tool — it tunes quality to hit a file-size budget, which this does not attempt.

What happens when the shapes don't match

Ask for a square from a 4:3 photo and something has to give. There are only three honest answers and this offers all three, because which one is right depends entirely on what you are uploading to.

A 4000 x 3000 photo asked for 1000 x 1000
Fit inside    1000 x 750     whole image kept, height falls short
Fill & crop   1000 x 1000    exact size, 500px trimmed from each side
Stretch       1000 x 1000    exact size, image visibly squashed

Fit inside never loses any of the picture, but you do not get the exact dimensions you typed — which fails a form that demands them. Fill and crop gives the exact size by taking a centred rectangle out of the middle, which is right for thumbnails and avatars and wrong when the subject is near an edge. Stretch gives the exact size by distorting, which looks bad and is occasionally the only thing a rigid upload form will accept.

The line under the fields shows the actual output dimensions and warns when cropping will occur, so the trade is visible before you commit rather than discovered in the download.

Leaving one field blank scales proportionally, which is the common case and avoids the question entirely: "1200 wide" keeps the shape and works everywhere.

Doing it on your own device

The images that go through resizers are, again, not usually abstract. They are a headshot for a job application, a photograph of a passport sized for a visa portal, a product shot for a listing, a screenshot of a document, a picture of a child for a school system.

Every one of those goes to a server if you use a typical online resizer, and the original — full resolution, complete with the GPS coordinates recorded by the camera — is what gets uploaded. The resized copy you download is smaller and stripped; the file you handed over was not.

Canvas has been able to do this in the browser for over a decade, so the server round-trip buys nothing but cost for the operator and exposure for you. Here the image is read from your disk, drawn by your own GPU or CPU, and written back locally. Check it in the Network tab, or disconnect from Wi-Fi and resize anyway.

A useful side effect: because the output is drawn from decoded pixels rather than copied, camera metadata including location does not come across. If you want to see what was in there before it went, the EXIF viewer linked below will show you.

Quality, formats and one thing to avoid

Downscaling is the safe direction. Removing pixels loses information you were not displaying anyway, and the result looks identical at the size it will actually be viewed. Upscaling is not: enlarging a 400-pixel image to 1600 invents nothing, it just makes each original pixel four times larger and softer. If a form demands dimensions larger than your source, the honest answer is usually to find a bigger original.

Choose the output format to match the content, not the input. Photographs belong in JPEG; screenshots, logos and anything containing text belong in PNG, where hard edges stay crisp. WebP is smaller than both at equivalent quality and is supported everywhere current, which makes it the best choice when the destination accepts it.

Converting to JPEG flattens transparency to white, since JPEG has no alpha channel. Keep PNG or WebP if the transparency matters.

And resize once, from the original. Each resize-and-re-encode cycle costs a little quality permanently, so going 4000 → 800 → 1200 is visibly worse than going 4000 → 1200 directly. Keep your originals and always start from them.

Frequently asked questions

How do I resize an image to exact dimensions?

Enter both a width and a height, then choose Fill & crop — that guarantees the exact size by taking a centred rectangle from the middle. Fit inside keeps the whole picture but one dimension will come out smaller than you asked, which some upload forms reject.

What is the difference between fit, fill and stretch?

Fit inside keeps the entire image and may come out smaller than the target. Fill & crop hits the exact size by trimming the overflow from the edges. Stretch hits the exact size by distorting the picture. The line under the fields tells you the real output size and warns when cropping will happen.

Should I use this or the image compressor?

Use this when a specific pixel size is required — a form, a thumbnail, a print spec. Use the compressor when the requirement is a file-size limit, because it tunes quality to hit a budget rather than taking dimensions literally.

Can I make a small image bigger?

You can, but it will look soft. Enlarging invents no detail — it just makes each original pixel larger. If something demands dimensions bigger than your source, finding a larger original is almost always the better answer.

Is my image uploaded?

No. It is read from your disk and redrawn in your browser, so it never leaves your device. That matters because the originals people resize are headshots, ID documents and product photos, usually still carrying the GPS coordinates of where they were taken.

Why does my resized photo look worse than expected?

Most often because it has been resized more than once. Every resize-and-re-encode cycle costs quality permanently, so 4000 → 800 → 1200 is visibly worse than 4000 → 1200 in one step. Always start from the original file.

Related tools