About
Why this exists
Nearly every free tool for a small file job asks you to upload the file first. People do it dozens of times a day with tax returns, passports, medical letters, signed contracts, salary slips and client work — handing a private document to a server they know nothing about, to perform a task their own computer could have done in a second.
The upload was never technically necessary. Browsers have been able to do this work for years, using the same capabilities that run video calls and photo editors. OpenTools is 567 everyday tools across 18 categories, built on that observation and nothing cleverer: the work happens in the tab you already have open, so there is no server to trust, no queue to wait in, and nothing to delete afterwards.
How it actually works
Your file is read into the page’s memory and processed there. Document work runs in a background worker so the page stays responsive; image work uses the browser’s own drawing engine; the background remover runs a compact AI model compiled to WebAssembly; audio and video trims copy the existing compressed samples rather than re-encoding them, which is why they finish almost instantly and lose no quality. Checksums use the browser’s built-in cryptography.
Two consequences follow, and both are the point. Your document never travels, so there is nothing to intercept or retain. And there is no round trip, so the work finishes at the speed of your own machine rather than the speed of someone’s queue. The measured evidence for the first is on the proof page; the enforcement behind it is on the security page.
Things this site will not do
Stated here so you can hold the project to them, and so that breaking one is visibly a broken promise rather than a quiet product update.
No upload of your work
Tools run in the tab. A version of this site that sent your file somewhere would not be a new feature; it would be a different product.
No account, ever
Nothing here needs to know who you are, so nothing here asks.
No advertising, tracking or session replay
Not as a setting you can switch off — as code that is not present, and a test that fails if it appears.
No paid tier, and no gated result
Support is voluntary and changes nothing about what you can do. A result is never held back.
No watermark or branding in your file
What comes out is your document, not an advertisement for this site.
Who runs it, and how it is paid for
This is a small independent project, not a company. There are no investors to satisfy and no growth target that would one day make advertising or an upload step look reasonable. It is not free because a paid tier subsidises it; it is free because it costs little to run when the visitor’s own computer does the work.
What it does cost — a domain and edge hosting — is covered by voluntary support from people who found it useful. Nothing is withheld from anyone who does not contribute, and the support page is deliberately kept apart from the tools: it never receives a file name, a job or a result.
The entire site is open source under the MIT licence. You can read every line, run it on your own machine, host it inside your own network, fork it, or take a piece of it for something else. If this project ever stops being maintained, nothing you rely on disappears with it — which is the practical reason the licence matters more than any promise on this page.
How this site is built
A rule the project holds to, because breaking it is the ordinary way a site like this becomes untrustworthy: no claim ships unless code or a test proves it. Tool counts come from the catalogue rather than from memory. A tool is listed only when the page behind the link genuinely runs it. Speed and privacy statements are measured, dated and tied to the build they were measured on, and when a claim cannot be supported it is removed rather than rephrased — which is how several confident-sounding lines came off this site rather than onto it.
If you find something on any page here that is not true, that is worth reporting, and the security page explains where to send it.