Privacy
The short version
Your files never leave your device. Every conversion and every PDF tool on this site runs inside your browser — a WebAssembly build of ffmpeg, pdf-lib, the browser’s own image codecs, or a text-recognition model that downloads and runs in the page. There is no upload step and no server that could receive your file.
You do not have to take that on trust. Open your browser’s network panel before you convert and watch: the file never appears in a request. That is a property of how the site is built, not a promise about how we behave with data we are given. We are not given it.
Two things on these pages come from another company, and both are about the page, not your files: Google Analytics, which counts visits, and Google AdSense, which shows the ads that pay for the site. Neither is given your file, its name or anything in it — their requests are in the same network panel, and none of them carries your file. Both can set cookies; in the EEA, the UK and Switzerland you are asked before they do.
1. The files you convert
They are read by the page, held in memory while the conversion runs, and handed back to you as a download. Nothing about them is transmitted, logged or retained. The same is true of file names, page counts and any metadata inside the file.
Where a conversion needs a large component — the ffmpeg core, or the text-recognition model — that component is downloaded to you. The direction is worth noting: the code travels to your file, not the other way round.
Scanned PDFs and text recognition
When a PDF has no text layer, the pages are recognised on your own machine by PP-OCRv5 mobile, running through onnxruntime in the browser. The model files are served from this domain and cached by your browser the ordinary way, so the download happens once. Neither the page images nor the recognised text is sent anywhere.
If that ever changes
Some formats genuinely cannot be processed in a browser. If a conversion is ever added that needs a server, the page for it will say so before you choose a file, in the badge at the top — the same badge that currently reads "Runs on your device" on every page. A change of that kind would be a change to the product, announced on the page, not a quiet edit to this policy.
2. What is stored in your browser
Our own: at most one cookie, named locale, holding the two-letter code of a language you picked from the language menu, so that your next visit opens in that language. It is set only when you pick one, holds nothing else, and goes to no site but this one. Until you pick, a visitor arriving from outside at a page with no language in its address is sent to the version in the language their browser asks for — or, failing that, the one for the country the connection comes from; that is decided afresh on each visit and nothing about it is stored.
Google’s: Google Analytics sets cookies (named _ga and _ga_ followed by an id) to tell a returning visitor from a new one, and Google AdSense sets cookies to limit how often an ad is repeated, to detect fraud and — only where you allow it — to show ads based on your interests. In the EEA, the UK and Switzerland none of these is set until you agree in the consent message shown on your first visit, and you can change that choice later from the same message.
The site’s own code writes nothing else: no localStorage, no IndexedDB, no service worker, no Cache Storage. The downloaded engines and models sit in your browser’s ordinary HTTP cache, which you control from your browser settings like any other site’s.
3. What is collected
Visit statistics, through Google Analytics: which pages are viewed, how a visitor arrived, the country and city the connection comes from, the browser, device and language, and how long the visit lasted. We use it to see which converters people use and which pages fail them. It receives no file, no file name and nothing typed into a tool. There is no tag manager, no session recorder and no other analytics tool.
Advertising, through Google AdSense, which chooses and shows the ads on these pages. To do that Google receives the address of the page, your IP address, browser details and its own cookies, and where you have allowed it, uses them to personalise ads. How Google uses this is set out at policies.google.com/technologies/partner-sites; personalised ads can be turned off at adssettings.google.com.
Server logs are the exception, and they are the ordinary kind: whoever hosts this site records requests — the IP address they came from, the time, the page and the browser string — because that is how a web server is operated and how abuse is dealt with. Those logs are about requests for pages, not about the contents of anything you convert, which never reaches them.
4. What is never collected
No account, so no name, email address or password. No payment details, because there is nothing to pay for. No contact list, no precise location, and no profile of what you convert: the analytics and advertising scripts see which page you are on, never which file you open on it.
Nothing is sold. Visit statistics and ad requests go to Google as described above, and nowhere else.
5. Your rights
Because your files are never collected, what can be requested, corrected or deleted is limited to two things: the host’s access logs and the visit statistics in Google Analytics. Data Google holds for advertising is governed by Google’s own tools, at adssettings.google.com and myaccount.google.com.
If you are in the UK, EU, or anywhere with statutory data rights, those rights apply to that data, and the contact address below is where to exercise them. You can clear every cookie and everything stored locally yourself at any time by clearing site data for this domain.
6. Children
The site is not directed at children, and nothing is knowingly collected from them. Since there is no sign-up, there is nothing for anyone to submit.
7. Changes
If what the site collects changes again — another tool, or any processing that moves off your device — this page will change with it and the date at the top will move. The badge on each converter page is the other place you would see it.
8. If you send us a file to report a problem
This is the one case where a file of yours reaches us, and it happens only because you decided to send it. It is worth describing precisely, because everything above says the opposite.
If a conversion fails or produces something wrong, a message to the address below is what gets it fixed. What helps most: which page you were on, which browser and version, what you expected and what you got. The technical detail in the grey panel under a failed conversion is the single most useful thing to paste — it names the component that gave up.
Take the private parts out first
A file that reproduces the fault helps enormously, but only send one if you can do it safely. Before you attach anything, remove what you would not want a stranger to read: crop the pages that matter, blank the names, numbers and addresses, or — usually the best option — rebuild the fault in a file you made up. A converter bug is a bug in how a shape is read, not in what the shape says, so an invented document with the same structure normally reproduces it exactly.
If a file cannot be made safe, do not send it. Describe it instead — how many pages, whether it is a scan, which application produced it, roughly how large. That is often enough.
What happens to a file you do send
It is used to reproduce the fault and for nothing else. It is not added to a dataset, not used to train anything, and not shared. It is deleted once the problem is fixed or closed, and you can ask for it to be deleted sooner at any point — a one-line message is enough, and you do not have to give a reason.
9. Who is responsible, and how to ask
Questions about this policy, or about anything the site does with data:
Contact
- Responsible for this site:
- Convertube
- Contact:
- convertubenet@outlook.com
- Feedback:
- Report problems and request features on GitHub. Issues there are public; for a file with anything private in it, email us instead.