About CocoPDF

Why it exists

Most PDF problems have a thirty-second fix, and most people lose an afternoon to them anyway. Part of that is the format being genuinely confusing. A scan looks exactly like a document until you try to search it. A rotated page looks fixed until you send it to somebody else. The rest of it is the software: install this, make an account, hit a daily limit, find a watermark across the page you needed for a client.

So the tools here are built to be finished in one visit. Open the page that matches your problem, drop the file in, get the result. If a setting matters it is on the page rather than buried in preferences, and the defaults are set to the answer most people need.

What actually processes your files

Sites like this one rarely say what is under the hood, which makes it hard to judge whether the output will be any good. CocoPDF runs established open source engines, each doing the job it is best at:

LibreOffice

Word, Excel, and PowerPoint conversions in both directions.

Ghostscript

Compression and repair, plus the PDF rendering behind image export.

ImageMagick

Everything involving images, including JPG and PNG output.

Tesseract

OCR, with trained data for English, French, Spanish, German, and Arabic.

pdf-lib

Page operations: merging, splitting, rotating, extracting, and page numbers.

All of it runs server-side on our own machines rather than being passed to a third party API, and rather than running in browser JavaScript, which is why a large conversion finishes in seconds instead of freezing the tab. The one deliberate exception is the document scanner, where edge detection and perspective correction happen on your own device. Photographs of passports, contracts, and bank statements never leave your machine unless you choose to build a PDF from them.

What happens to what you upload

Your file is written to a temporary directory, processed in an isolated container, and deleted the moment your download finishes. That is not a scheduled job you have to trust: the deletion is attached to the response, so it also runs if you close the tab partway through or lose your connection.

If something goes wrong during processing, a corrupt file or a conversion that times out, the upload is deleted immediately as part of handling the error. A sweeper runs every few minutes on top of that and removes anything on disk older than an hour, so nothing survives a crash or a restart. Uploads are capped at 50 MB each.

Nobody reads your documents. Nothing is scanned for content, indexed, kept as a backup, or passed to anyone else. There are no accounts, so there is no file history to accumulate and nothing tying a document to you in the first place. The Privacy Policy sets this out in the form a compliance team will want.

What we will not do

No watermarks on output, at any quality, on any tool. No daily limit on how many files you process. No account wall in front of a result you have already waited for. Your documents are not training data and are not sold to anyone. If any of that ever has to change, it will be written on this page rather than discovered by a user halfway through a job.

Who runs it

CocoPDF is built and run by Oussama EL, its Founder and CEO. It is a small operation rather than a company with a support department, which is the reason the tools stay free and also the reason the person answering contact@cocopdf.com is the same person who wrote the code. If a tool is doing something strange, that message goes straight to someone who can fix it.

Oussama EL

Founder and CEO, CocoPDF

View on LinkedIn

Getting in touch

Bug reports are genuinely useful, particularly the ones that include the tool, your browser, and what the file was like. Describe the document rather than attaching it, since we would rather not hold copies of anyone's files in an inbox. Email contact@cocopdf.com or use the contact form. Replies usually take a day or two.