Starting the machine
Loading the editor. Your files stay on this device.
Merge, reorder, rotate, resize, sign and annotate PDFs, then export with no quality loss. Your files are opened and edited inside this browser tab and are never sent to a server. There is no signup, no account and no cost. Add a PDF or an image above to start, with nothing to install.
There is no upload endpoint. The page reads your file into memory using the browser's own file APIs, every edit happens in this tab, and the exported PDF is assembled here and saved straight back to your device. Nothing is transmitted, so there is nothing for anyone to store, read or leak. It is not a retention policy that could change later: there is no data.
The site does count how often each action is used, and only that. Each tracked action sends one event name from a fixed list of eight and nothing else, so no filenames, no annotation text and no image data ever leave the page. There is no third-party analytics script, no cookie and no persistent identifier, which is why there is no consent banner. Nothing is stored on your device either, so closing the tab discards everything.
Display and export are separate systems. Previews are drawn with pdf.js purely so you can see the pages, and those pixels are never exported. The export is built with pdf-lib at the PDF object level: page content is copied byte for byte, rotation is stored as the page's rotation metadata rather than re-rendering, and embedded JPEGs keep their original bytes so they are never re-compressed.
A 300 DPI scan exports at 300 DPI.
Most free web PDF tools work by uploading your document to the operator's servers, processing it there, and giving back a download link. That means your file sits on someone else's infrastructure, subject to their retention window, their staff access and their breach risk. The free tier is usually gated by an email signup, a daily quota, or a watermark on the output, and many of them rasterize the document, which quietly degrades scans and text.
Free PDF Machine never uploads the file, asks for nothing, adds no watermark, and copies the original page objects instead of re-rendering them. Those differences are verifiable rather than promised: there is no upload step to opt out of, because there is no server-side processing to upload to.
Everything runs on your own hardware, so very large documents are bounded by your device's memory. There is no OCR, no PDF form filling, no compression targeting a smaller file size, and no collaboration or cloud storage, because those either need a server or are out of scope. Nothing is kept between sessions. A PDF locked with a password to view cannot be opened. If you need any of those, another tool is the right answer.
Yes, completely. There is no paid tier, no trial period, no watermark on the exported PDF, no page or file-size limit imposed by the product, and no feature reserved for a premium plan. It does not ask for a credit card and there is nothing to cancel. The price is 0 USD and the whole tool is available to everyone who opens the page.
No. Your files are opened and edited inside your own browser tab and are never sent to a server. There is no upload step in the product because there is no server-side processing to upload to: the page reads the file from your device into memory, and the edited PDF is assembled in the browser and saved straight back to your device. Because nothing is transmitted, there is nothing for anyone to store, read or leak.
No. There is no account system at all. You do not enter an email address, you do not verify anything, and there is no login screen. You open the page and start working. The consequence is that nothing is saved between visits: when you close the tab, the working document and any signatures you drew are gone, so download your exported PDF before you leave.
It can merge several PDFs and images into one document, reorder pages by dragging them, rotate pages in 90 degree steps, delete pages, resize pages to A4, US Letter or the document's dominant size, add a hand-drawn signature, add text boxes in a choice of sizes and colours, place images onto pages, highlight selected text, and draw free-hand highlighter marks over any page including scanned ones. You can act on many pages at once with multi-select, and export either the whole document or just the pages you selected.
Partly. The page itself has to be loaded over the internet at least once, because there is no installed app and no offline cache. But once the page has loaded, the editing work is fully local: opening files, every edit, and building the exported PDF all happen on your device with no further network activity. If your connection drops mid-session, you can still finish your document and export it.
No. The exported PDF is rebuilt from the original bytes of the file you opened rather than from what you see on screen. Page content is copied at the PDF object level, so text stays as text and vectors stay as vectors; rotation is recorded as page metadata instead of re-rendering the page; and embedded JPEGs keep their original bytes, so they are never re-compressed. A high-resolution scan comes out at the same resolution it went in.
It is designed for exactly that case. Contracts, medical records, bank statements and identity documents stay on your machine, which is the difference between this and an online PDF service that requires you to upload the file first. The one thing to be aware of is that the tool cannot open a PDF that is locked with a password to view.
Free PDF Machine is made by Amit Raz at RZ AI Labs, an AI consulting and software development practice based in Haifa, Israel. There is a project page for Free PDF Machine and a write-up of how it was built with Claude Code. The source is public on GitHub.