About

pdfpix is a set of PDF tools that run inside your browser tab. Nothing is uploaded and no account is needed. Every tool reads its own output back with a second, independent PDF reader and compares it with your original file, so a result that cannot be verified is never offered to you as a download.

The idea

Most free PDF tools ask you to upload the file. That is a strange thing to do with a bank statement, a contract or a passport scan. Browsers can do this work now: they have a PDF reader, a PDF writer, a JPEG encoder and a font subsetter, all of them open source and all of them fast enough. So the file stays with you.

The rule

A tool that cannot check its own work is a guess. Every pdfpix tool reads its output back with an independent reader and compares it against the source. If the check fails, the candidate is dropped, not shipped. The one that matters most: a document with selectable text comes out with selectable text, or it does not come out at all.

Who builds it

pdfpix is built by Rasty Turek, a security engineer out of Google, through his company Wraxle LLC. He knows what happens to a document once you hand it to somebody else's server, and he did not want that to be the price of compressing a PDF.

Keeping the file costs you nothing here. Compression, redaction, merging and conversion all run at full strength on your own machine, so the private choice is also the capable one.

Write to support@pdfpix.com. It reaches the person who wrote the tool you were using.

Where it stands

All 31 tools are built and running. No page here claims a feature that does not run. The tool hub has a card for every one of them; this is the same set, sorted by the job.

What is not here

No account is needed, and there is no watermark on the output, no daily file limit and no queue. There is also no server doing the work, which sets the edge of what this site can offer. Jobs that need a real machine, such as reading text out of a scan or running the same operation over two hundred files at once, are not browser jobs and are not pretended to be. When a tool cannot do something, its page says so in plain words instead of failing quietly.

If something goes wrong

Every error message a tool can show is written out on the support page, with what it means and what to try next. The code behind the tools is open source, and the licenses page names each library and its license. The how it works page covers the checks themselves, and the privacy page covers what the site does and does not record.

Where to go next