How img-tools works — in-browser image processing explained

The short version

When you use a tool on img-tools, your image is loaded into your browser's memory and processed by code running on your own device — using the same web technologies that power modern web apps. The result is generated locally and offered to you as a download. At no point is the image sent across the internet.

The technologies that make it possible

Canvas API: The HTML5 Canvas lets JavaScript read and write the individual pixels of an image. Operations such as resizing, cropping, rotating, drawing watermarks, and re-encoding to JPEG, PNG, or WebP are performed directly on a canvas in your browser tab.

WebGL: For real-time, GPU-accelerated effects (filters, adjustments, previews), img-tools uses WebGL, which runs shader programs on your device's graphics hardware. This keeps editing fast and smooth without any server round-trip.

WebAssembly (WASM): Heavier work — AI background removal, upscaling, object detection, and some format decoders — runs as WebAssembly modules. WASM is a fast, sandboxed binary format that executes inside the browser at near-native speed, again entirely on your device.

Cache API: AI tools download their model files once and store them in your browser's cache. After the first use the model loads from local storage, so the tools work offline and never need to re-download.

Why 'no upload' is true by design

A traditional online editor is a thin front-end that sends your file to a backend server, processes it there, and returns the result. img-tools has no such backend for image processing. The processing code is shipped to your browser and runs there, so there is simply no server that receives your image.

This is a property of the architecture, not a promise in a policy. Even our privacy-friendly analytics only ever counts anonymous page views — it never touches your image data. The site's strict Content-Security-Policy further restricts where the page may send data, and all fonts and AI models are self-hosted on the same domain.

How to verify it yourself

You don't have to take our word for it. Open your browser's developer tools (F12 or right-click → Inspect), switch to the Network tab, and use any tool — compress an image, remove a background, whatever you like. You will see the page load its code and (for AI tools) a one-time model file, but you will never see your image being uploaded.

It is not just a promise — it is enforced in code. Every tool page installs a runtime network guard that intercepts fetch, XMLHttpRequest, and sendBeacon and refuses any request carrying your file data to another server. You can confirm the guard is active by typing window.__imgToolsNoUpload in the developer console — it returns true. So even a bug could not leak your image.

You can also disconnect from the internet after the page has loaded: the instant tools keep working, and AI tools keep working once their model is cached — proof that the processing is happening on your device.

What this means for you

Because everything runs locally, img-tools is private by default, works offline, has no file-size limits imposed by an upload pipeline, adds no waiting time for round-trips, and never adds watermarks to your results. It is suitable for confidential, sensitive, or simply personal images that you would rather not hand to a third-party server.

And it is 100% free, with no account and no paid tier — because we are not in the business of storing or monetising your files.