Image Resizer

About this tool

Upload forms and social platforms have strict pixel limits — a 4032×3024 phone photo almost never fits without resizing. Enter a width, a height, or both, and the image is resized on our server by Sharp, whose default Lanczos3 filter keeps downscaled photos crisp. Fill in only one dimension and the other is calculated from the original proportions. Fill in both and the Fit mode decides what happens: Contain fits the whole image inside the box and pads the rest, Cover fills the box and crops the overflow from the centre, Fill stretches to the exact size, Inside keeps the image within both limits, and Outside makes it at least as large as both. Images are never enlarged beyond their original size, so small sources are not blown up into blurry copies. The output keeps your original format (JPEG stays JPEG, PNG stays PNG, WebP stays WebP). Files up to 20 MB are accepted and deleted from the server within seconds of the result being returned.

Why use it

• Exact Pixel Dimensions: Type the width and height a platform demands, such as 1200×630 for an Open Graph image, and get exactly that box back. Whole-pixel values are validated before upload so typos are caught early.

• Proportions Kept Automatically: Leave either Width or Height empty and Sharp calculates the other from the source ratio. This prevents the accidental squishing or stretching that is the most common resize mistake on portrait photos.

• Lanczos3 Resampling: Sharp uses Lanczos3 as its default downsample filter. It preserves high-frequency detail better than bilinear or bicubic resampling, meaning downscaled photos look noticeably sharper — especially when reducing resolution by more than half.

• Five Fit Modes: Cover fills the box and crops overflow, ideal for square thumbnails. Contain fits the whole image and pads the gaps. Fill stretches to exact dimensions. Inside and Outside treat your numbers as maximum or minimum bounds while keeping the ratio.

• No Accidental Upscaling: The resizer never enlarges an image past its original resolution, so requesting 4000 px from a 1000 px source will not produce a soft, interpolated file.

• Files Auto-Deleted Server-Side: Uploaded originals and the resized output are stored in a temporary directory during processing and deleted within seconds of the download response completing. Nothing is retained in storage.

How to use

1. Upload a JPEG, PNG, or WebP image up to 20 MB
2. Enter a target Width and/or Height in pixels; leave one blank to keep the original proportions
3. If you set both, choose a Fit mode: Contain, Cover, Fill, Inside or Outside
4. Click Resize Image and wait for the server to return the result
5. Check the preview and the new file size, then click Download

When it helps

• Before uploading product images to Shopify or WooCommerce, which have explicit pixel dimension requirements
• When preparing Open Graph images for social sharing at the standard 1200×630 pixels
• When Slack, GitHub, or a project management tool rejects an image because it exceeds their upload dimension limit
• When creating profile avatars that must fit a strict 256×256, 400×400, or similar square requirement
• Before submitting photos to a print service that specifies exact pixel dimensions at 300 DPI
• When generating multiple size variants of a hero image for responsive srcset attributes

Examples

Tips

• Always downscale rather than upscale — start with the highest-resolution source you have, since this tool will not enlarge a small image anyway.
• If the target box is square but your source is landscape, choose Cover — it keeps the centre of the image and trims the edges, so frame your subject in the middle first.
• For Open Graph images, 1200×630 is the accepted size that Twitter, LinkedIn, and Facebook all display without re-cropping or adding black bars.
• Leave one dimension empty when you only care about width (or height); the other is worked out from the original ratio.
• Need a smaller file as well as smaller dimensions? Run the result through the Image Compressor or convert it to WebP with the Image Format Converter.

Frequently Asked Questions

Glossary

Aspect ratio

The proportional relationship between an image's width and height, expressed as a ratio such as 16:9 or 4:3. Maintaining aspect ratio during resize prevents the image from appearing stretched or squished.

Lanczos3 filter

A high-quality image resampling algorithm that uses a sinc-based kernel to reconstruct pixel values during scaling. It preserves high-frequency detail (sharp edges and fine texture) better than bilinear or bicubic filters.

Cover vs contain fit

Cover scales the image to fill the target box entirely and crops any overflow, guaranteeing no empty space. Contain scales the image to fit entirely within the box and fills remaining space with padding or background colour.

Pixel density (DPR)

Device Pixel Ratio — the number of physical screen pixels per CSS pixel. A 2× retina display shows images at double resolution, so a 400×400 image on a retina screen only fills a 200×200 CSS pixel area at full quality.

Letterboxing

Padding added to the top and bottom (or sides) of an image to fill a target aspect ratio without cropping or stretching the original. Named after the black bars on widescreen films shown on standard-definition televisions.

Resampling

The process of calculating new pixel values when an image is scaled to a different resolution. The resampling algorithm determines whether the result looks sharp or blurry at the new size.