How local processing works
When you choose a file on PaperCircuit, it is read by code that already sits in your browser tab. The file is never sent to us, so there is nothing for us to keep, scan or leak.
What happens when you press “Create”
- Your browser reads the file.The file picker hands the bytes to the page’s JavaScript. They stay in this tab’s memory.
- An engine does the work on your CPU.pdf-lib, pdf.js, qpdf or tesseract.js — all downloaded from this site the first time you use a tool — process the bytes locally.
- The result is built as a local download.The new PDF is a
blob:URL created on your device. Closing the tab frees it.
The engines
| Engine | Used for | License |
|---|---|---|
| pdf-lib | Merging, splitting, page picking, rotation, blank pages, N-up, overlays, page numbers, watermarks, annotations, signatures, metadata, images and text to PDF. | MIT |
| pdf.js | Reading and rendering pages: thumbnails, previews, PDF to images, text extraction, compare, and the page images used by grayscale, redaction and OCR. | Apache-2.0 |
| qpdf (WebAssembly) | Linearizing for fast web view, rebuilding damaged files, AES-256 encryption and password removal. | Apache-2.0 |
| tesseract.js | Optical character recognition for scanned pages; the English model is served from this site and cached by your browser. | Apache-2.0 |
| fflate · jsdiff | Zip downloads and image decoding; the word-level text diff in Compare PDFs. | MIT / BSD-3 |
All of them are open-source and served from ilovepdfpdf24.com itself. No ad network or tracking script is part of the site. The site is delivered through Cloudflare, which adds its cookieless page-load beacon (static.cloudflareinsights.com); it reports page timing and visit counts and has no access to the files you open.
Check it yourself in 30 seconds
- Open any tool, for example Compress PDF.
- Open your browser’s developer tools (
F12or⌥⌘I) and switch to the Network tab. - Choose a PDF and run the tool. You will see the engine files load (
.js,.wasm), but no request that carries your document. You can even switch the browser to offline after the page has loaded — most tools keep working.
Passing files between tools and the Workbench
“Continue with this result” stores the file in your browser’s IndexedDB under this site, opens the next tool, and deletes the stored copy as soon as that tool picks it up. The Workbench keeps every intermediate file in memory only.
What the server does see
Like any website, our web server receives normal page requests: your IP address, the page URL, browser user-agent and time. These access logs are used for security and capacity only and are rotated. They never contain file names or file contents. See the privacy policy for details.
Limits of doing it locally
- Speed depends on your device. Very large files, OCR and image-heavy compression are slower on older phones.
- Memory is limited by the browser tab; files of several hundred MB may fail on mobile.
- Some conversions (Word, Excel, PowerPoint, PDF/A) need engines that can’t yet run privately in a browser, so we don’t offer them rather than upload your file.