What this tool does
Almost every other PDF question starts with a fact about the file that nobody has told you: how many pages it has, how big the page is, which version of the format it claims, and whether the text on it is real text or a photograph of text. This tool answers all of that at once, and it is the page to read before deciding what to do with a PDF someone sent you.
How it works
A PDF describes its own pages in points, not pixels and not millimetres. One point is exactly one seventy-second of an inch, so an A4 page is 595.276 by 841.890 points, and a page claiming 595 by 842 is a rounded value that a print shop can measure the difference on. This tool reports points, then millimetres, then names the size when it matches a standard one, because "A4" is an answer and "595.28 by 841.89" is only a measurement.
The distinction people most often get wrong is between a page that is landscape and a page that is rotated. A landscape scan has a media box that is genuinely wider than it is tall. A portrait scan that a scanner or a phone saved sideways has a tall media box plus a rotation flag attached, so the media box still says portrait. Those two files look identical on screen and need opposite fixes, which is why rotating a sideways scan once sometimes appears to do nothing.
The version comes from the header line, and it is a claim rather than a guarantee. A file headed 1.7 may use nothing from 1.7, and a 1.4 file may have been written by software newer than any reader you are likely to open. It is reported because people search for it, and labelled as what it is. Alongside it, the tool counts the revision markers in the file: every time a PDF is saved again it is appended to rather than rewritten, so a document saved six times carries six revisions.
Font embedding is checked by walking the resource dictionaries of every page and comparing the fonts named there against the list of fourteen that every PDF reader is required to have. Anything outside that list has to be embedded, and a font that is named but not embedded is the usual cause of text that displays correctly on the machine that made the PDF and as empty boxes everywhere else. This is the single most useful diagnostic on the page for a file that will not print correctly on someone else's machine.
Worked example
A supplier has sent a 60-page quotation as a PDF. Before forwarding it internally you need the real page count, the actual paper size, and whether the prices are selectable text or a scan.
- Drop the file in. Nothing is uploaded, so a supplier's confidential pricing stays in the tab
- Read the page count and confirm the page size is named A4 rather than a rounded figure
- Check the rotation column: a page listed as rotated is a portrait page with a flag, not a landscape one
- Check the font report for anything not embedded, which is what would break it on the recipient's machine
You know the file is 60 pages of A4 with all fonts embedded, so it will render identically on any machine, and you can tell whether the prices are searchable text before trying to copy them.
Accuracy and limitations
- A password-protected file cannot be opened here, so a locked PDF reports as unreadable rather than reporting its contents. Unlock it first.
- The version number is read from the file header and describes what the writer claimed, not what the file actually uses. A 1.4 file can contain 1.5 features and still open in older readers by luck.
- Font embedding is reported per font across the document, not per page, so a document with one unembedded font reports that font once rather than flagging each page it appears on.
- This tool reads the file. It does not change it, so nothing here can damage a document.
Frequently asked questions
- How do I count the pages in a PDF?
- Open the file in this tool and the page count appears immediately, along with the size of each page in points and millimetres. The count comes from the page tree in the file itself rather than from a filename or an estimate, so it is the number a reader would show.
- What does it mean if a PDF says the fonts are not embedded?
- The document names a font that is not built into the file and is not one of the fourteen standard fonts that every reader has. The text will usually look right on the computer that created the PDF and appear as blank boxes or a substitute typeface on someone else's. Ask for an export with fonts embedded, or print to PDF from a desktop application rather than exporting from a web app.
- Why does my PDF say A4 is 595 by 842 points and not 600 by 800?
- Because A4 is defined in millimetres, not in round numbers. It is 210 by 297 millimetres, and at 72 points to the inch that is 595.276 by 841.890 points. The 595 by 842 figure is a rounding of a real measurement, and printers care about the difference even though most screen viewers do not.
- Is a rotated page the same as a landscape page?
- No, and the distinction matters. A landscape page has a media box that is wider than it is tall, so the content was laid out sideways. A rotated page has an upright, taller media box with a rotation flag attached, so the content is upright and the reader is being told to turn it. They look the same on screen and need different corrections.
- Does this tool upload my PDF anywhere?
- No. The file is read into your browser's memory, inspected there, and discarded when the tab closes. You can confirm it by opening your browser's network panel while it runs and seeing that no request is made. That matters for a file containing a supplier's pricing or a client's bank statement.
- Why can't it read a password-protected PDF?
- The library this runs on cannot open an encrypted file at all, rather than merely not knowing the password. Unlocking a PDF is a separate operation that we do not perform here, so a protected file is reported as unreadable instead of reporting partial information.