Guide
Why your images should not be uploaded
Most “free online image compressor” sites work the same way: you upload a file, a server compresses it, you download the result. It is convenient, invisible, and for a lot of images it is a bad trade. Here is what that upload actually costs, and how to tell whether a tool is really doing the work on your machine.
The file leaves your control
The moment a file is uploaded it exists on hardware you do not own, under a retention policy you probably have not read. A passport scan for a visa form, a medical letter, a photo of a contract, an unreleased product shot: these are the files people compress most often, because forms always have size limits. Many services delete uploads after an hour; some keep them for weeks; some keep them indefinitely and reserve the right to use them to improve their service.
Uploads create accounts, logs and subpoenas
Even a tool that deletes immediately has already logged your IP address, the timestamp, the file size and often the filename — which may contain your name or your client’s. That metadata is exactly what gets requested in a data access request, and it exists for every file you have ever compressed. Local processing simply does not generate it.
Third parties are usually in the chain
A typical compressor is a front end in front of cloud storage, a queue, a worker and a CDN. Each hop is another vendor with its own access rules. You are trusting all of them for a task your own laptop can do in under a second.
What local processing actually means
In Lighterpix the encoders — MozJPEG, OxiPNG, libwebp, libavif, libjxl — are compiled to WebAssembly and run inside a Web Worker in your tab. The file you drop is read with the File API, decoded into memory, re-encoded, and offered back as a download. There is no upload endpoint to call, because there is no server doing the work.
How to verify it yourself
Open the browser developer tools, switch to the Network tab, clear it, and compress an image. A local tool shows requests for the page and its codecs on first load, then nothing while the file is processed — no POST, no multipart upload. The stronger test is to switch the network off after the page has loaded: a local tool keeps working, an upload-based one stops.
Questions
Can a website download my files without me knowing?
Any page you give a file to can read it, including ones that process it locally. The difference is that local processing does not need to transmit it anywhere, and you can verify that in the network panel.
Is offline use a fair test?
It is the simplest one that is hard to fake. Load the page, turn off your network, then compress an image. If it still works, the processing was local.
When is uploading unavoidable?
When the work needs hardware you do not have — video transcoding at scale, for example. For still images, resizing and re-encoding a few megabytes is well within a browser’s reach.