How in-browser file tools keep your files private

Step 1: the browser reads the file you choose

Web pages can't look around your disk. The File API, the part of the browser that handles files, only gives a page access to files the user makes available: through a file picker, or by dragging and dropping them onto the page. The page then reads the contents of that file into the device's memory.

Reading is not uploading. Up to this point nothing has been sent anywhere; the data sits in the browser tab, the same way a photo you open in an app sits in that app.

Step 2: the work runs on your device

Compressing a JPEG or building a PDF needs serious code. Browsers can run WebAssembly, a low-level format that MDN describes as bringing near-native performance to the web. It serves as a compilation target for languages such as C, C++ and Rust, so proven image libraries can run directly in the browser alongside JavaScript.

That is how our tools work: MozJPEG for JPG, OxiPNG for PNG, libwebp for WebP and libavif for AVIF all run in background workers inside your browser tab. PDFs are handled with pdf-lib and PDF.js, which are also code that runs on the device. The speed you see depends on your phone or computer, not on a server queue.

Step 3: the result is saved from memory

When the work is done, the new file exists only in the tab's memory. The page turns it into a link with URL.createObjectURL, which creates a blob URL pointing to that data. Pressing Download saves it from there to your device. The address of such a link starts with blob: and refers to memory in your own browser, not to a server.

How to check it yourself

Two simple tests, from easy to thorough.

The airplane mode test: open the tool page and let it load fully. Switch on airplane mode or disconnect from Wi-Fi. Now process a file. If it still works, the file can't have been sent anywhere, because there was no connection to send it over.

The Network panel test, on a computer: in Chrome, press Control+Shift+J on Windows or Linux, or Command+Option+J on a Mac, then click the Network tab. Other desktop browsers have a similar panel. DevTools only logs activity while it's open, so open it first, then process a file and watch the list.

What the page still loads, and what that means

An in-browser tool is still a website. Opening it downloads its own code: the HTML, scripts, WebAssembly files and fonts. Pages that show ads or count visits also load those services, and they learn that someone visited the page, as on any website. None of that carries the contents of your file.

Once a page has been opened, ours keeps a copy so it works offline, which is why the airplane mode test works even on the second visit.

How this differs from upload-based converters

With an upload-based converter, your file travels to someone else's computer. What happens next, how long it's kept, who can access it and whether it is logged, depends on that company's systems and policies, which you can't inspect. For a holiday photo that may not matter. For a passport scan, a contract or medical letter, it's a reasonable thing to avoid.

The honest limits

A website can only control what happens inside its own page. A browser extension with permission to read the pages you visit can see what the page sees, and malware on the device can see anything. Neither is something any website can prevent. Downloads also land in your normal downloads folder, so on a shared computer, delete them when you're done.