Image Resizer

An image resizer to resize an image or photo by pixels or percentage — change image size with aspect-ratio locking and presets for Instagram and other common social and display sizes.

Loading Image Resizer…

Runs entirely in your browser. Files stay on your device — nothing is uploaded. Verify it yourself

Does resizing an image reduce its quality?

Making an image smaller is clean — the browser interpolates well and the result stays sharp. Making it larger cannot add detail that was never captured, so an enlarged image is always softer than the original.

Cost: Free, no limits
Upload: None — runs on your device
Sign-up: Not required
Works offline: Yes, once loaded

How to resize a photo

Choose an image
Select the image you want to resize. It is processed on a canvas in your browser, so nothing is uploaded.
Set the dimensions
Enter a width or height, use a percentage button, or pick a preset for a common social or display size. Leave the aspect-ratio lock on unless you have already cropped to the target shape.
Choose format and quality
PNG preserves transparency and is lossless; JPEG and WebP let you trade quality for size.
Download
Check the preview and download. Note that camera metadata, including any GPS coordinates, is not carried over.

Examples

A 4000×3000 photo at 25%1000×750
Percentage scaling keeps the aspect ratio automatically.
A 4000×3000 photo to 1200px wide1200×900 with the ratio locked
The height follows so the image is not stretched.

Key features of the Image Resizer

Resize by pixel dimensions or by percentage
Aspect-ratio lock so images do not distort
Presets for common social and display sizes
Runs on your device — nothing is uploaded

Supported formats

Input

JPG
PNG
WebP
GIF
BMP
AVIF

Output

JPEG
PNG
WebP

When should you use this?

Fitting an image to a maximum dimension for a website.
Preparing a photo for a profile or banner slot.
Reducing a very large image so it uploads or emails successfully.
Making a set of images consistent in size.

Limitations

Enlarging an image cannot add detail that was never captured. Upscaling interpolates between existing pixels and produces a softer result — no resampling method changes that.
Unlocking the aspect ratio lets you distort the image. That is occasionally what you want and usually not.
Resizing a lossy image means re-encoding it, so a small amount of additional quality is lost even at high settings.
Very large source images are bounded by browser memory, since the work happens on your device.

Troubleshooting

The resized image looks soft

If you enlarged it, that is inherent — there is no additional detail to show. If you reduced it, check you are not viewing it scaled up in the preview.

The image looks stretched

The aspect-ratio lock is off and the width and height no longer match the original proportions.

Nothing happens on a very large file

Browser memory. Close other tabs and retry, or reduce the image in two steps.

Common mistakes to avoid

Upscaling a small image and expecting it to look sharp. The detail is not there to recover.
Turning off the aspect ratio lock by accident and stretching everything.
Resizing repeatedly through several rounds instead of going back to the original each time.

Pro tips

Tips for better results

Source file
Resize down from the original, not from a reduced copy
Resize down from the original rather than from an already-reduced copy. Each resize is a re-encode.
Screens
Size for the largest display size, plus a little for high-density screens
For screens, work out the largest size the image will actually be displayed at and add a little for high-density displays. Anything beyond that is wasted bandwidth.
Order of work
Halving the dimensions removes three quarters of the pixels
Resize before compressing. Halving the dimensions removes three quarters of the pixels, which saves far more than any quality setting.

Concepts behind image files

The terms that decide what changes when an image is compressed, resized or converted.

Resolution: The pixel dimensions of the image or frame, such as 1920×1080. It sets the amount of real detail present. Enlarging afterwards interpolates between existing pixels and produces a softer picture — no amount of upscaling recovers detail that was never captured. Reference: CSS Values and Units Module Level 4, W3C
Colour depth: How many bits describe each pixel's colour. Most formats use 8 bits per channel, giving about 16.7 million colours. Indexed formats such as GIF store a palette of at most 256 instead, which is why photographs converted to them show visible banding. Reference: GIF89a Specification, CompuServe, hosted by W3C
Metadata: Information stored alongside the content itself — camera settings, GPS coordinates, artist and album tags, author names. Conversion frequently drops it, which is usually a privacy benefit and occasionally a loss if you relied on it.
Colour space: The rule that maps the numbers in a file to visible colours. sRGB is what browsers assume, Display P3 and Adobe RGB describe wider ranges used in photography, and CMYK describes ink rather than light. A file converted without its profile keeps its numbers and changes its colours. Reference: sRGB, the default colour space, ICC
DPI and PPI: Dots or pixels per inch: a number describing how large the picture is meant to print, stored as a tag beside the pixels. It changes nothing on a screen, which draws by pixel count alone, and nothing about the detail present — only the size a printer lays the same pixels down at. Reference: CSS Values and Units Module Level 4, W3C
EXIF data: The block cameras and phones write into a photograph: exposure, lens, orientation, the date, and on a phone usually the coordinates it was taken at. Converting writes a fresh file without it, which loses the shooting data and removes the location along with it. Reference: CIPA DC-008: Exchangeable image file format (Exif), CIPA
Pixel: One addressable square of a raster image, holding a single colour. A 1920×1080 picture is 1920 across and 1080 down, about 2.07 million in all, and that count is fixed when the file is written — which is why enlarging one later invents the pixels it does not have. Reference: PNG Specification, Third Edition, W3C

Standards and terms behind the Image Resizer

The named things this tool works with, and the documents that define them.

Canvas APISpecification: The HTML drawing surface that lets a browser decode, resize and re-encode an image without a server. Reference: Canvas API
ResamplingHow it works: Working out the new pixels when an image changes size; enlarging invents detail that was never captured.
Aspect ratioHow it works: The proportion of width to height; changing it without cropping stretches the picture.
Social media uploadWhere it is used: Platform upload rules, which usually accept a short list of formats and re-encode whatever arrives.
E-commerce listingsWhere it is used: Marketplace image rules — fixed dimensions, white backgrounds and a short list of accepted formats.

Also searched as: resize a photo to exact pixels, change image dimensions, profile picture size, resize for instagram.

Why aspect ratio matters

Changing width and height independently stretches the image. The lock keeps the original proportions by adjusting the other dimension automatically, which is what you want unless you are deliberately fitting a fixed frame.

When you do need an exact size that does not match the original ratio — a social banner, say — resizing alone will distort it. Crop to the target ratio first, then resize.

Downscaling is safe, upscaling is not

Reducing an image discards pixels, and modern browsers do it with good interpolation, so the result looks clean. Enlarging has to invent pixels that were never captured. No amount of interpolation adds detail — it just spreads the existing detail over more pixels, giving a soft, slightly blurry result. The tool warns when you cross that line.

AI upscalers do better by predicting plausible detail, but they are inventing it, which makes them unsuitable where accuracy matters.

Resize for the size you display at

Serving a 4000px photo into a 800px slot wastes bandwidth on every visitor and slows the page for no visual gain. Resize to roughly the display size — or twice it, for high-density screens — and the page gets faster with no visible difference.

Metadata

The resized image is drawn onto a fresh canvas, which means EXIF data — camera model, settings, and any GPS coordinates — is not carried over. For sharing photos publicly that is usually what you want. If you need the metadata preserved, keep the original alongside.

Privacy and file handling

The Image Resizer runs entirely in your browser. Your file is read by the page, processed on your own device, and the result is handed straight back to you — no copy is sent anywhere, because none needs to be.

You can confirm this rather than take our word for it: open your browser’s developer tools, watch the Network tab, and use the tool. No request carries your file or your input. It also keeps working with the connection off.

Details are in the privacy policy.

What next?

After resizing, compression trims what is left.

Compress the result

Image Resizer — questions

Can I resize a photo on my iPhone or Android phone?

Yes, in the phone’s browser: there is nothing to install, and the photo stays on the phone. On an iPhone or iPad, choose JPEG or PNG as the format — Safari cannot write WebP.

Does resizing reduce quality?

Making an image smaller is clean — the browser interpolates well and the result looks sharp. Making it larger cannot add detail that was never captured, so the result is softer than the original.

What is aspect ratio locking?

It keeps the original proportions: change the width and the height follows automatically. Turn it off only when you need an exact size and have already cropped to that ratio, or the image will stretch.

What size should images on my website be?

Roughly the size they are displayed at, or twice that for high-density screens. Serving a 4000px photo into an 800px slot costs every visitor bandwidth for no visible benefit.

Is EXIF data preserved?

No. The image is redrawn on a canvas, so camera metadata and any GPS coordinates are stripped. That is usually desirable when sharing photos publicly; keep the original if you need the metadata.

Part of Image Tools · Browse every tool