Image Resizer
Resize images to exact pixel dimensions with an optional locked aspect ratio.
Drop an image here or click to browse
Supports JPG, PNG, WEBP
All resizing happens in your browser. Your image is never uploaded.
Resizing an image to exact pixel dimensions is one of the most routine tasks in web and design work, and one of the easiest to get subtly wrong. Change only the width and the image stretches; guess at the height and you distort it; upscale a small image and it turns soft and blurry. The key facts are that shrinking is nearly always safe while enlarging is not, and that keeping the aspect ratio locked is what prevents distortion. This tool resizes to precise dimensions with an optional locked ratio, entirely in your browser, so your images never leave your device.
How the resizer works
Upload an image and enter your target width or height. With the aspect ratio locked — the usual choice — entering one dimension sets the other automatically to keep the image undistorted; unlock it to force exact dimensions at the risk of stretching. The tool redraws the image at the new size through the Canvas API and lets you download the result. Everything happens locally; the image is never uploaded to a server.
The rules of resizing
Locked ratio (safe):
new height = new width × (orig height ÷ orig width)
Downscaling: 1000px → 400px ✓ stays sharp
Upscaling: 400px → 1000px ✗ turns blurry
Display ratios (CSS):
1× normal 2× retina/HiDPI needs 2× the pixels
Distortion happens only when ratio is unlocked.The most important rule is that you can shrink an image well but cannot truly enlarge one. Downscaling discards pixels the image already has, so a 2000-pixel photo scaled to 800 stays crisp. Upscaling has to invent pixels that were never captured, and no amount of interpolation adds real detail — it just blurs or, with AI tools, guesses. This is why the workflow is always to start from the largest source available and scale down to each size you need, never to scale a small image up.
What to know about resizing
- 1Downscale freely, but never upscale for quality. Shrinking a large image to a smaller size looks great because it is throwing away surplus detail; enlarging a small image cannot add detail that was never captured, so it goes soft. Always resize down from the biggest original you have, and if you only have a small image, accept its size rather than blowing it up.
- 2Resize to the actual display size to save huge amounts of weight. A 3000-pixel photo shown in an 800-pixel slot is downloading nearly four times the pixels it needs, wasting bandwidth and slowing the page. Matching the pixel dimensions to the display dimensions is often the single biggest file-size win available, bigger than compression.
- 3Account for retina screens by resizing to 2× the display size. High-density displays pack two device pixels into each CSS pixel, so an image shown at 400 CSS pixels looks sharpest exported at 800 actual pixels. The balance is serving enough resolution for retina without the bloat of an unnecessarily huge file — 2× is the usual sweet spot.
- 4Keep the aspect ratio locked unless you deliberately want to distort. Unlocking the ratio lets you hit exact width and height, but any mismatch with the original stretches or squashes the image. For almost every use, lock the ratio and enter just one dimension; reach for unlocked only when you genuinely need to force a non-native shape and accept the distortion.
- 5Resize before compressing, and keep an original master. Resizing to the display size first means compression then works on far fewer pixels, compounding the savings. And because downscaling discards detail permanently, always resize from a full-resolution master rather than from an already-shrunk copy, so you can regenerate any size you need later.
Frequently asked questions
Why does my image look blurry after resizing?
Almost certainly because you enlarged it. Upscaling a small image forces the software to invent pixels that were never captured, and no interpolation can add real detail — the result is soft or blurry. Downscaling, by contrast, stays sharp because it discards surplus pixels. The fix is to start from a larger original and scale down to the size you need; if you only have a small image, there is no way to genuinely enlarge it without losing sharpness.
How do I resize without distorting the image?
Keep the aspect ratio locked and change only one dimension, letting the tool calculate the other proportionally. Distortion happens only when width and height are scaled by different amounts, which stretches or squashes the content. With the ratio locked, entering a target width sets the matching height automatically, so the image keeps its shape. Unlock the ratio only when you deliberately need exact dimensions and accept the resulting distortion.
What size should I resize images to for the web?
Match the pixel dimensions to how large the image will actually display, then double that for high-density screens. An image shown in an 800-pixel-wide slot should be around 800 pixels for standard screens or 1600 for retina sharpness. Serving a 3000-pixel image into an 800-pixel space wastes most of the download. Resizing to the display size — the biggest easy win for page speed — should come before compression, not after.
Can I make a small image larger without losing quality?
Not really. Enlarging an image asks the software to create detail that was never in the original, and standard resizing can only interpolate, which blurs. AI upscaling tools can plausibly invent detail and sometimes look impressive, but they are guessing, not recovering real information. The reliable approach is always to source or keep a large original and scale down; a small image cannot be truly restored to a larger sharp one.
Is my image uploaded when I resize it?
No. The resizing runs entirely in your browser through the Canvas API, so the image is processed on your own device and never transmitted, logged, or stored. You can confirm this by disconnecting from the internet; the tool keeps working because there is no server involved. This keeps private photos, client work, and unpublished images on your machine throughout.