FAQ & Privacy
What file types can I convert?
In: .md .txt .doc .docx, up to 4MB. Out: PDF .docx .md
What is Markdown output for?
Turning a Word document into plain text you can keep in a notes app or a repository. Headings, lists, tables, links and emphasis all carry over. Images can't live inside a Markdown file, so each one is noted in place instead. Hebrew, Arabic and other scripts export exactly as they are, though Markdown itself has no notion of text direction, so whether a Hebrew file reads right-to-left is up to the app you open it in.
Is there a length limit?
Around 300 pages, alongside the 4 MB file limit - longer than almost anything anyone actually converts. The two limits are separate because file size is a poor guide to length: text doesn't compress on upload, so a 4 MB text file is over a thousand pages, while a 3 MB Word file full of images is usually about thirty. If a document is too long we say so up front and tell you roughly how many pages it is, instead of failing at the last moment. For .md and .txt, Convert in my browser has no length limit at all, because it runs on your machine.
How does Markdown work?
Headings, lists, tables, emphasis, code blocks and links become real formatting. We detect Markdown saved as .txt or .docx too, and you can force it on or off in Options.
Why is .doc worse than .docx?
Legacy .doc (Word 97-2003) is a binary format with no maintained open reader that reproduces formatting. We can recover the text reliably, but not fonts, tables or images. If you can, open the file in Word and save it as .docx - that path keeps headings, lists, tables, images, footnotes and links.
What is “Convert in my browser”?
For .md and .txt to PDF, the conversion runs entirely on your device and nothing is uploaded. Your PDF comes from your browser's own Save as PDF, so you finish in the print dialog. Word files need the server to unpack them.
Do you keep my files?
No. The file is held in memory only for as long as the conversion takes, and is gone when the response is sent. It is never written to disk, never copied anywhere, and there is no database, storage bucket or queue in the system to put it in.
Do you log what I convert?
No. The conversion endpoint writes no log lines at all - not the filename, not the file size, not the type, not whether it worked. In fact your filename never reaches us: the browser sends only the file's bytes and names the download itself.
Is there any tracking?
Only privacy-preserving page counts (Vercel Web Analytics), which set no cookies and store no persistent identifier. That is why there is no cookie banner to dismiss. Nothing about your document is ever sent to it.
Why doesn't my PDF look exactly like the original?
Word documents carry a lot of information that has no exact equivalent outside Word. We preserve headings, lists, tables, images, footnotes, links, bold/italic/underline, highlighting, alignment, indentation and relative text sizes. Text colour is currently not preserved. If something specific looks wrong, tell us - that's how this list gets shorter.
Does it handle right-to-left and special-character languages?
Yes. Fonts covering Latin, Greek, Cyrillic, Hebrew, Arabic, Devanagari, Thai, Chinese (Simplified and Traditional), Japanese and Korean are bundled with the service, and right-to-left documents are laid out right-to-left. Nothing is fetched from Google Fonts or any other third party.
Is there a limit to how many files I can convert?
No daily cap. There's a soft burst limit to stop automated abuse, which normal use won't reach.
Is it really free?
Yes. No registration, no payment, no ads, no upsell.
How it actually works
Your browser sends the file's bytes over HTTPS. The server reads them into memory, extracts the content, lays it out with a bundled stylesheet and fonts, renders it with a headless browser, strips the metadata from the result, and sends the PDF back. Then the memory is released. There is no step in which anything is stored.
- Your filename never reaches us. Only the file's contents are sent.
- We don't log conversions. Not the file, not its size, not its type.
- The renderer has no network access. Every request it makes is blocked except the bundled fonts, so a document that links to an external image cannot make our server fetch it.
- Output metadata is stripped, including the timestamp the file would otherwise record. This applies to both PDF and Word output.
- Word comments and document properties are dropped, so the author, company and revision history inside a .docx can't leak into what you share.
- Image metadata is stripped, including EXIF GPS coordinates, camera serial numbers and capture times. A photo taken at home and pasted into a report would otherwise carry your address into the PDF.
- The renderer can't write caches to disk, so the browser engine can't leave your document behind in a scratch file either.
What we can't promise
Your file does travel to a server, and it is encrypted in transit rather than end-to-end. We necessarily see the contents in memory in order to convert them.
We also can't speak for our host. The site runs on Vercel, and their platform keeps its own short-lived request records - timestamps, IP addresses, response codes - the same as any web host, entirely outside our code. We never write your document, its name or anything about it into those. Removing that layer altogether would mean not using a hosting provider, which isn't something a free service can do.
What we can say is that nothing is retained by us and nothing about your conversion is recorded - and that you don't have to take our word for it. Open your browser's developer tools and watch the Network tab while you convert; with Convert in my browser ticked, you will see your file go nowhere at all. Every response also carries a Content-Security-Policy that forbids this site from sending data to any other domain, and that is enforced by your browser. See the what we connect to section to learn more.