How to compress images without uploading them anywhere

Almost every popular image compressor works the same way: you drop in a file, it travels to a server somewhere, gets compressed there, and comes back smaller. It works, and for a single screenshot it is fine. But that round trip carries two costs most people never think about — until they matter.

The privacy cost of uploading

When you upload an image to compress it, you are handing a copy of that file to someone else's computer. For a holiday photo, no big deal. For a client's unreleased product shots, a medical scan, a legal exhibit, or a screenshot with personal data in it, that copy is a liability. You are trusting a third party's retention policy, their security, and their word that the file is deleted afterward.

Plenty of professionals work under agreements that simply do not allow this. A designer under NDA cannot send pre-launch assets to an unknown server. The safest file is the one that never leaves your device in the first place.

The speed cost of uploading

Uploading is also slow in a way that scales badly. One image is quick. Two hundred images means two hundred uploads, two hundred server round trips, and two hundred downloads — bounded by your upload bandwidth, which is usually the slowest part of a home or office connection. A batch that should take seconds can stretch into a long, babysit-the-progress-bar chore.

Compression can happen in the browser instead

Modern browsers can run the same compression codecs that used to require a server or a desktop app, thanks to WebAssembly — a way to run high-performance, compiled code inside a web page. The encoders for WebP, JPEG, and AVIF all run this way. That means the entire job — reading the file, compressing it, writing the result — can happen on your own machine, with no upload at all.

The difference is easy to verify. Open your browser's developer tools, switch to the network tab, and compress an image. With a local tool, you will see no file leaving your machine. It is not a privacy promise written in a policy; it is an architectural fact you can watch.

What you give up, and what you don't

Running locally means the work uses your computer's processor and memory rather than a data center's. On a normal laptop that is plenty for large batches. On a phone or a restricted browser it is more limited, which is why local tools usually cap batch sizes on mobile to avoid crashes. In exchange, you get privacy by default, no upload wait, and no per-file limits imposed by someone else's free tier.

Try it

Velacompress is built entirely around this idea. Drag in a folder, choose WebP, JPEG, or AVIF, and everything is compressed in your browser — no upload, no size limits, no subscription. Open the app and watch the network tab while you do it.