Skip to content
ToolDock
All guides

Guide · 4 min read · updated 2026-10-19

What "runs in your browser" actually means for your files

The difference between browser-local tools and cloud tools, what each one sends over the network, and how to tell which is which before you use it.

ToolDock is free and shows ads. Google and its partners may use cookies to serve and measure them.Privacy Policy

Two architectures, one page

Online file tools come in two shapes. Server-side tools upload your file, process it on a remote machine, and hand you back a download. Browser-local tools ship the processing code to you — the conversion, compression or generation happens on your own device, and the file never leaves it. Both are legitimate; they suit different jobs and different sensitivities.

ToolDock labels every tool with where it runs, in plain language, before you use it: a green "on your device" badge for local tools, a blue "secure cloud" badge for the ones that need the network.

What a local tool can and cannot send

A genuinely local tool has no file upload step: your document is read with the browser's file API, processed with JavaScript and WebAssembly, and the result is saved from memory to your downloads. You can verify the claim with your browser's network tab — while a local tool processes a 50 MB video, no request carrying that video appears. What can still flow normally: loading the page itself, its fonts and code, and anonymous feature analytics.

The limits of local processing are physical, not policy. Everything runs on your device's CPU, so very large batches take real time, and anything needing a heavy model (background removal, large-scale OCR language packs) depends on what fits in a browser tab.

When the cloud version is the right call

Some jobs are only possible with a server: moving a file between two devices, letting a colleague upload into your inbox, or anything that must happen while you are offline. ToolDock's transfer tools work that way, and they are built accordingly: files sit in private storage, access is gated by a random transfer code, downloads go through short-lived signed URLs that expire, and sessions are purged automatically when they age out. Abuse controls — rate limits and lockouts on code guessing — run server-side where they have to.

The habit worth building takes ten seconds: before using any online tool with a file you would not hand to a stranger, check the processing-location label. Choose local for anything sensitive, and use cloud tools for the jobs that genuinely need connectivity — knowing exactly what each one does with your file.

Keep reading