About

About Compress Images

Compress Images is a browser tool run by Marlow Ridge Software. It works on your file inside your browser, so the file stays on your device.

Run by
Marlow Ridge Software
Contact
support@compress-images.online

What the site does

Compress Images compresses and converts JPEG, PNG, WebP and AVIF images. You drop one file or a batch of them on the home page, you choose the output format, a quality and a size, and each file is decoded, resized and encoded on your own device. Every finished file comes back with its size before and after, a preview of the result and its own download link, and a run can be stopped at any point without leaving a half-written file behind.

Nothing about your file is sent to us. The whole run happens in your browser, which is what how it works describes in detail and what the security page shows the evidence for.

What the tool does not do

Compress Images does not read HEIC, does not read or write JPEG XL, and has no zip download for a batch. An input of more than 40,000,000 pixels, or larger than 32 MiB, is refused with a message that names the limit, and the rest of the batch still finishes. Those are fixed numbers rather than a reading of your machine, so a different computer does not raise them: the file has to change. A lossy format cannot restore detail the encoder threw away, so a smaller file is not always a better one.

What runs on your device, and what comes with it

The tool is built from open-source WebAssembly codecs, each published by someone other than this site and each under its own license, which is not the same as the license of the package that wraps it. The credits page names every module, its version and the license text this site serves for it, and it says plainly where its own evidence runs out.

Who runs it

Compress Images is run by Marlow Ridge Software. Questions, corrections and problem reports go to support@compress-images.online. For a security problem, see the security page.

How the claims on these pages are checked

Each claim these pages make about your file and about what your browser stores is listed with the code or the test that shows it, and the site’s automated checks fail when that proof goes missing. Numbers are either cited or worked out from a stated model and marked as illustrative. Where the evidence runs out, the page says so instead of filling the gap: the JPEG codec’s package records no upstream version, and the PNG codec’s readme and its shipped license notice disagree. If you find something wrong, write to us and we will correct the page.