QuickMergeOn-device resizing — the selected image is not sent to a QuickMerge processing server.
Resize Image Online — Increase or Reduce Picture Size
Short answer: To resize an image, enter the target pixel dimensions or a percentage, preserve the aspect ratio when needed, then export it. QuickMerge resizes images in your browser without uploading them.
Enlarge a picture or shrink it: enter exact pixel dimensions or a percentage, keep the aspect ratio locked, and download JPG, PNG or WebP. The whole resize happens on your own device.
- Free to use
- No sign-up
- Exact pixel entry
- Runs in your browser
- No uploads
How do I resize an image to exact dimensions?
Drag & drop here
or click to browse
Resize Photos and Images to Any Size
Whether you need an avatar at exactly 400×400, a banner at a specific width, or a photo shrunk to upload faster, this tool resizes JPG, PNG, and WebP images to precise pixel dimensions. Tick Keep aspect ratio and set just the width (or height) and the other dimension is calculated for you, so nothing gets stretched.
Resizing is done on an HTML canvas in your browser. Everything runs on your device via open-source browser libraries — your file is never uploaded. You can verify it yourself.
Next steps
Related tools
- change the framing, not just the size — when the aspect ratio is wrong.
- convert a JFIF before resizing — if the file came off a camera or an email attachment.
What to do with the result
- compress after resizing — the two together get you under most upload limits.
- remove EXIF data — before posting the resized image.
Browse all image tools, or see every QuickMerge tool.
Wondering how big a file this can take? We measured it: see Resize image on the measured limits page, or the browser support matrix.
Frequently Asked Questions
Load the image, untick nothing, and type a larger width — the height follows automatically while Keep aspect ratio is on — then download. Enlarging cannot add detail the original never captured, so treat roughly 200% of the source width as the point beyond which softening becomes obvious on screen.
It can. Resizing up spreads the pixels you have over a bigger grid and the browser interpolates the gaps, so fine detail softens. Starting from the largest original you have, rather than from a copy that has already been shrunk or forwarded through a chat app, makes far more difference than any setting here.
No. The image is read from your device, resized on a canvas in the page and downloaded from the page. Nothing is transmitted, and there is no account to create.
Leave Keep aspect ratio ticked and set only one dimension — the other is computed automatically to keep the proportions correct.
Yes. Pick JPG, PNG, or WebP as the output, or keep the original format. WebP usually gives the smallest file at the same quality.
Making an image smaller is clean. Enlarging beyond the original size will look softer, since the tool cannot invent detail that was not captured.
No — resizing happens entirely in your browser. The image never leaves your device.
Set the pixel dimensions first, then lower the quality slider if the file is still too large. Dimensions dominate file size, so halving the width typically cuts the file to roughly a quarter before quality is touched at all. For a hard byte target, compress the image after resizing.
Enter the exact pixel dimensions the authority publishes, and untick Keep aspect ratio only if they also specify an exact crop. If the source photo is the wrong shape, crop it to the right proportions first, then resize — cropping then resizing preserves the subject, resizing then cropping does not.
This tool works one image at a time so you can confirm each result. For a batch where every file needs the same treatment, resize them in sequence — nothing is uploaded, so there is no per-file wait.
Re-encoding usually drops most EXIF, but do not rely on that as a privacy measure. If the goal is stripping location and camera data, use the metadata remover, which is built for it and tells you what it found.
Match the largest slot it will occupy, then double it for high-density screens. A banner spanning a 1200 CSS-pixel container wants roughly 2400 pixels of source width; a thumbnail in a 300-pixel card wants about 600. Anything beyond that is bandwidth with no visible return.
Making an image smaller is safe — you are discarding pixels you no longer need and the result stays sharp. Enlarging is where quality goes, because no tool can recover detail the sensor never recorded; beyond about 200% of the original width, softening becomes obvious. If the file is a JPG, resize the original rather than a copy that has been re-saved several times.
Reduce the pixel dimensions first — halving the width typically cuts the file to roughly a quarter — and only then lower the quality. Dimensions dominate file size far more than the compression setting does. For a hard byte target, compress the image against it after resizing.
Decide the print resolution first, then multiply: 6 × 4 inches at 300 DPI is 1800 × 1200 pixels, which is what you enter. Use 300 DPI for photographic printing and 150 for a poster read from a distance.
Not by entering mismatched dimensions with the ratio unlocked — that stretches the picture and it shows immediately on faces and straight lines. Crop to the ratio you need first, then resize the cropped image to the exact pixel size.
The same way as any other image: enter the dimensions and export. Keeping PNG as the output format preserves transparency and re-encodes losslessly. If the picture is a photograph rather than a graphic, exporting as JPG or WebP instead will produce a far smaller file.
Yes, one after another with the same entered dimensions, which keeps every output identical. Since nothing is uploaded there is no waiting between files.
Pixels determine detail; file size is only the consequence
Resizing changes pixel dimensions. Compression changes how efficiently those pixels are stored. Decide the required dimensions first and optimise quality second.
Work backwards from display size
A 1200-pixel-wide image is normally sufficient for a 600 CSS-pixel slot on a high-density screen. Supplying a 5000-pixel camera original wastes bandwidth without adding visible detail. For email or forms, follow the portal’s stated maximum dimensions rather than guessing from megabytes alone.
Avoid accidental distortion
Keep aspect ratio locked unless the destination specifies an exact crop. Forcing a 4:3 image into a square by changing width and height independently stretches faces and circular objects. Crop to the target shape first, then resize the cropped result.
Upscaling has limits
Making a small image larger creates new pixels by interpolation; it cannot recover missing texture or readable text. Modest enlargement can be useful for layout, but screenshots and scanned documents should be captured again at higher resolution when accuracy matters.
Resolution guidance updated 11 July 2026 · Review methodology
The three things that make a resized image look worse than it should
Resizing itself is lossless arithmetic. Almost every complaint about a "blurry resize" traces back to one of these instead.
Re-saving a JPG repeatedly
Every JPG save re-applies lossy compression. Resizing a JPG that has already been saved four or five times compounds artefacts that were baked in long before this tool touched it. If you still have the original, resize that rather than the copy that has been round-tripped through chat apps.
Upscaling past the captured detail
Enlarging cannot recover information the sensor never recorded. A 400-pixel-wide image stretched to 1600 will look soft no matter which tool does it. As a rough ceiling, going beyond about 200% of the original width is where softening becomes obvious on screen.
Fighting the destination’s own resize
Many upload forms resize again on their side. If a portal states a maximum dimension, match it exactly — supplying something larger means your image is resized twice, and the second pass is one you do not control.
How to resize an image to the dimensions something is asking for
Almost every resize is driven by a requirement somewhere else — an upload limit, a print size, a screen resolution. Working backwards from that requirement is what makes it a single pass rather than three.
Pixels, inches and DPI are the same number wearing different hats
Screens work in pixels; print works in inches at a stated resolution. The bridge between them is DPI: an image to be printed at 6 × 4 inches at 300 DPI needs 1800 × 1200 pixels, and that is the whole calculation. Resizing a picture in inches means deciding the print resolution first — 300 DPI for photographs, 150 for something being read at arm's length — and multiplying. An image that is already smaller than that cannot be rescued by entering a larger number, because the detail was never captured.
Common target dimensions, and where they come from
1920 × 1080 is a full-HD screen and the safe default for a wallpaper or a presentation background. 1200 pixels wide covers most blog content areas; 2400 covers the same slot on a high-density display. Profile pictures are usually square at 400 × 400 or larger. When a service publishes a maximum, match it exactly — supplying something larger means it resizes again on its side, and that second pass is one you do not control.
Changing the aspect ratio distorts; cropping does not
Unlocking the aspect ratio and entering mismatched numbers stretches the image, and on anything containing a face or a straight line the result is immediately obvious. When the target shape differs from the source shape, crop to the right proportions first and then resize — that keeps the subject and the geometry intact. Resizing then cropping throws away detail you were about to keep.
Resizing a batch to one size
Where every file needs the same treatment, work through them in sequence with the same entered dimensions so every output matches. Nothing is uploaded, so there is no per-file wait between them — the only cost is your own clicks, and the results are consistent in a way that eyeballing each one is not.