Add a Watermark to PDF Free Online

Is this PDF watermark tool completely free?

Yes, completely free. There are no usage limits, no account required, and no secondary Convixy watermark is added to your output. The only mark on your PDF is exactly what you configure in the editor — this is different from many "free" watermarking tools that quietly stamp their own logo across your document unless you upgrade to a paid tier, or cap you to a handful of conversions per day.

There's also no artificial restriction on how many times you can use the tool, no daily quota, and no watermark strength or size limited to a "preview" version with a paid unlock. You can process one document or fifty in a row with the same settings and the same result each time.

The only limits that exist are technical, not commercial: a 15 MB upload limit on the source PDF, a 5 MB limit on the watermark image, and a 25 MB output cap. These exist to keep server processing fast and reliable for everyone, not to push you toward a paid plan. If your PDF is near the limit, run it through Compress PDF first — most documents shrink by 60–80% with no visible quality loss.

Does the watermark appear on every page?

Yes. The watermark is applied to every page at the same relative position, size, opacity and rotation — the tool reads each page's exact width and height and recalculates the placement individually, rather than using one fixed set of pixel coordinates for the whole document. That means a mixed document containing both A4 portrait pages and A4 landscape pages, or pages of genuinely different sizes, will still show the watermark in the correct relative spot — say, centred, or 30pt from the bottom-right corner — on every single page.

This also means the watermark text or image scales consistently with your chosen font size or size percentage regardless of the underlying page dimensions, so a "30% of page width" logo will look proportionally the same on a small page and a large one, rather than becoming disproportionately large or small.

There is currently no option to apply the watermark to a subset of pages (for example, only the cover page, or only pages 5 through 10) — it always applies uniformly across the whole document. To watermark only specific pages, use Split PDF to extract just those pages first, apply the watermark here, then reassemble the full document with Merge PDF.

What packages does this tool require on the server?

The watermarking engine uses three Python packages, each responsible for a distinct part of the pipeline. pypdf handles reading the source PDF's existing page objects and merging the generated watermark layer onto each one without touching the underlying content stream. reportlab is the canvas engine that actually draws the watermark — it renders text with real font metrics (so centring and rotation are pixel-accurate) or builds the transparent overlay page that an image gets placed onto. Pillow (PIL) is only invoked in image watermark mode, where it opens the uploaded JPG, PNG or WEBP, normalises it to RGBA, and applies the requested opacity to its alpha channel before handing it off to reportlab. All three are mature, widely used open-source packages available via pip, with no exotic system dependencies beyond what Pillow needs for image codecs (libjpeg and zlib, which are present on virtually every Linux distribution by default).

To install them on your server: pip3 install pypdf reportlab Pillow --break-system-packages

Then verify they are accessible to your web server user specifically — a package installed under your own shell user but not under www-data (or whichever user runs PHP-FPM/Apache) is a common source of the "module not found" errors seen in the debug output above: sudo -u www-data python3 -c "import pypdf, reportlab, PIL; print('OK')". If that command fails while the same import succeeds under your own user, the fix is almost always to install the packages system-wide rather than into a user-specific site-packages directory, or to point PHP's exec() call at the correct Python interpreter path.

Does watermarking reduce the quality of my PDF?

No. The watermark is generated as a completely separate ReportLab canvas layer — effectively a small, invisible-except-for-the-watermark PDF page — and merged onto your original page object using pypdf's page-level merging. Crucially, this merge operation works at the level of PDF page content streams, not pixels: your original page is never decoded into an image and re-encoded, which is exactly the kind of lossy round-trip that degrades quality in cruder watermarking tools. Existing body text stays fully searchable and selectable, embedded images retain their original compression and resolution exactly as they were, and any embedded fonts remain intact and unmodified.

Text watermarks are drawn as genuine vector outlines using the requested font at the requested size, so they stay perfectly sharp at any zoom level — printing a watermarked page at high resolution or zooming to 800% in a PDF viewer will not reveal any pixelation or blurring in the watermark text, the same as with any other vector text in a PDF.

Image watermarks are embedded at their source resolution and re-encoded only to apply the requested opacity to the alpha channel — if you upload a low-resolution logo, it will still look low-resolution when scaled up on the page, so starting with a reasonably high-resolution source image (particularly for a transparent PNG) gives the cleanest result. As for file size, the output is typically only a few kilobytes larger per page than the input for text watermarks, since a text-only overlay layer is extremely lightweight; image watermarks add a bit more depending on the image's own file size and how many pages it's stamped onto.

Can I watermark a password-protected PDF?

It depends on which kind of "password-protected" you mean, since PDFs actually support two distinct and unrelated protection mechanisms. A user password (sometimes called an "open password") encrypts the file itself and must be entered just to view the document at all — PDFs protected this way cannot be processed here, because the engine needs full, unencrypted read access to each page's content stream before it can even locate where to draw the watermark, and this tool intentionally does not attempt to guess, crack, or bypass that encryption. You'll need to remove the open password in your PDF software first (most PDF readers and editors that can open the file with the correct password can also re-save it without one), watermark the now-unencrypted file here, and then optionally re-apply password protection afterwards using your own PDF tool.

An owner password (or "permissions password") is different — it doesn't block opening the file at all, it just restricts specific actions like printing, copying text, or editing, while the content itself remains fully readable by any PDF library. PDFs protected only this way can usually be watermarked here without any issue, since the restriction is enforced by compliant PDF readers rather than by encrypting the content stream. If you're unsure which type your PDF has, the simplest test is whether you're prompted for a password just to open and view it in a normal PDF viewer — if not, it's very likely only owner-restricted and should work fine.

Are my files private and secure?

Yes. All transfers between your browser and our server happen over encrypted HTTPS, the same standard used for online banking, so the contents of your PDF and any watermark image can't be intercepted in transit. Once your watermarked PDF is downloaded, both the original uploaded PDF and any watermark image you provided are deleted — the server also runs a periodic cleanup pass that removes any leftover watermarked output files older than one hour, so nothing lingers indefinitely even if a download is interrupted or never completed. We do not open, inspect, index, or share the contents of your documents at any point in that process; the only interaction with your file is the automated watermarking step itself.

No account or sign-in is required to use this tool, which means there's no user profile, upload history, or persistent identifier tying a given document back to you. This makes it a reasonable choice for confidential reports, legal contracts, medical documents, financial statements and other sensitive material — that said, if your organisation operates under strict compliance requirements (HIPAA, GDPR data-processing agreements, or similar), you should confirm that using a third-party web tool for document processing is permitted under your specific policies before uploading regulated data anywhere online, including here.

How do I change or remove a watermark after applying it?

Once merged, the watermark becomes part of each page's actual content stream in exactly the same way as the rest of the page's original text and graphics — there is no separate, toggleable "watermark layer" preserved in the output file that could later be hidden or deleted independently. This is a deliberate trade-off: the same permanence that makes a watermark impossible for us to strip back out afterward is also what makes it meaningfully difficult for anyone else to casually remove it from a document you've shared, which is usually the entire point of adding one in the first place (marking a document as a draft, confidential, or property of a particular owner).

If you applied the wrong text, colour, size, or position, there is no in-place "edit" option — you'll need to go back to your original, unwatermarked PDF and run it through this tool again with the corrected configuration. For that reason, it's worth keeping a copy of the clean, unwatermarked original in your own files any time you watermark something, rather than treating the watermarked output as your working copy going forward.

Once you're happy with your watermarked PDF, a few natural next steps: use Compress PDF if you need to reduce its size before emailing or uploading it, Merge PDF to combine it with other watermarked or unwatermarked documents into a single file, or Split PDF if you only need to distribute a subset of its pages afterward.