The two kinds of online converter
Every web-based file tool falls into one of two architectures, and the difference matters more than any privacy policy either one publishes.
Server-side converters upload your file over the network, process it on a machine you do not control, store the result temporarily, and hand you a download link. This is how the large majority of free converters work, because it is far easier to build and it works on any device.
Client-side converters ship the conversion code to your browser instead, usually compiled to WebAssembly, and run it on your own hardware. The file is read from disk into browser memory and never leaves the tab. There is no upload, no server copy, and no retention window, because there is nothing to retain.
- Server-side: your file is transmitted, stored briefly, and deleted on the operator's schedule
- Client-side: your file is processed in the browser tab and never transmitted
- Neither approach is dishonest, but only one removes the need for trust
How to tell which one you are using, in 30 seconds
You do not have to take anyone's word for it. Open your browser's developer tools, switch to the Network tab, and run a conversion. If the file is being uploaded, you will see a request carrying roughly the file's size in the request body. If nothing of that size leaves, the work is happening locally.
Two weaker signals are worth knowing. A converter that works with your network disconnected after the page has loaded is necessarily local. And a converter that shows a queue, a progress bar tied to your connection speed, or a per-day conversion limit enforced across devices is almost certainly server-side.
- Open DevTools, Network tab, run the conversion, watch for an upload the size of your file
- Disconnect from the network after the page loads: if it still converts, it is local
- Queues, wait times and cross-device limits imply a server
When uploading is genuinely fine
Server-side processing is not a scandal. If you are converting a holiday photo, a public brochure or a meme, the confidentiality risk is essentially zero and it is not worth thinking about.
Some work also genuinely cannot happen locally. Optical character recognition on a large scanned archive, machine translation, and large language model summarisation all need compute or model weights that are impractical to ship to a browser. When a tool tells you plainly that it processes server-side and why, that is the honest version of the trade-off.
When you should not upload
The category that matters is files carrying obligations. Signed contracts, employee records, patient data, financial statements, unreleased designs, legal discovery material, anything under an NDA, and anything covered by GDPR or HIPAA all fall here.
In those cases the problem is not that a converter operator is likely to read your file. It is that uploading it may itself be the violation. Many data processing agreements require you to know and name every processor that touches personal data. A free converter you found through a search result is not on that list, and adding it retroactively is not a thing you can do.
- Contracts, HR files, medical and financial records
- Anything under NDA or covered by GDPR, HIPAA or similar
- Unreleased product designs and pre-announcement material
- Legal discovery documents
Questions worth asking of any converter
If you must use a server-side tool for confidential work, get answers to these before you upload. A tool that cannot answer them clearly has told you something.
- Is the file uploaded at all, and can I verify that in the Network tab?
- How long is it retained, and is deletion automatic or manual?
- Is it encrypted in transit and at rest?
- Are third-party subprocessors involved, and who are they?
- Is the content used to train models?
- Is there a data processing agreement available?
Where Filoven sits
Filoven runs most of its tools entirely in the browser. For those, there is no upload, which is why every one of them works with no account and no daily quota tied to a server.
A minority of tools genuinely need a server: OCR at scale, document translation, structured extraction, transcription and image upscaling. Those are labelled as server-processed on their own page, they require signing in, and they consume credits. We would rather tell you which is which than claim a blanket guarantee that a hosted tool cannot honour.
Common questions
Do online file converters keep your files?
Server-side converters keep your file at least long enough to process it and serve the result, typically deleting it after a stated window such as one or 24 hours. You cannot independently verify deletion. Client-side converters never receive the file, so there is nothing to keep.
Can I convert files without uploading them?
Yes. Converters built on WebAssembly run the conversion code inside your browser, so the file is read from your disk into the tab and processed locally. You can confirm no upload happens by watching the Network tab in developer tools while you convert.
Is it safe to convert a confidential PDF online?
Only with a converter that processes locally in your browser. For a signed contract, patient record or anything under NDA, uploading to a third-party server may itself breach your data processing obligations, regardless of how trustworthy the operator is.
Are paid converters safer than free ones?
Not inherently. Price signals a business model, not an architecture. A paid server-side tool still receives your file; a free client-side tool still does not. What matters is where the processing happens, which you can verify yourself.
How can I verify a converter does not upload my file?
Open developer tools, go to the Network tab, and run a conversion. An upload appears as a request whose body is roughly the size of your file. As a second check, disconnect from the network after the page has loaded - a local converter will still work.