Lee este artículo en español.

How to find out who really wrote a PDF (and change it)

Every PDF carries a set of details you don't see when you read it, but you do when you check the document's "properties": who's listed as the author, what software created it, when, and sometimes a title that has nothing to do with what you're actually reading. That's metadata, and it travels with the file whether anyone looks at it or not.

It turns into a real problem in cases like:

• Someone sends you a PDF and you want to know who actually wrote it, not just what the filename says.
• You're about to send a document to a client and the title still reads "Template 2024," or the author is someone who no longer works with you.
• You need to strip out a company name or a previous author before sharing a file outside.
• You're generating PDFs automatically and want them all to carry the same title and keywords.

PDF metadata handles both sides of that: reading what a PDF already has, or changing only what you want.

Check what metadata a PDF has right now

Send the PDF with no metadata fields and the tool responds in read mode: it gives you back the metadata object exactly as it's stored in the file right now. That's the step for checking who's really listed as the author before assuming anything, or for seeing what title and dates a PDF someone sent you is carrying.

Change only the fields you care about

As soon as you include at least one metadata field in the request (even an empty one, to clear that specific field), the tool switches to write mode. It only touches the fields you give it; everything else in the PDF, including the rest of the metadata, stays as it was.

The available fields are:

• title, author, subject, creator, producer: plain text.
• keywords: you can send it as comma-separated text or as an array. In read mode it always comes back as an array, but that's an approximation: internally a PDF only stores a single keywords string, so it's not an exact conversion.
• creation_date and modification_date: ISO 8601 format, like 2026-08-06T10:00:00Z. An invalid format returns a 400 error.

Output format

In write mode, the response is either the binary PDF or a JSON body with the PDF in base64, depending on whether you request output=pdf or output=json. Read mode is always JSON: there's no PDF to return, just the data.

Every call, whether reading or writing, costs 1 credit.

FAQ

Does changing the title in the metadata change what shows up on the pages? No. The title, author and the rest of the metadata are separate from the visible content. Changing them doesn't touch a single character of what's written in the PDF.

Why do the keywords come back as an array when I sent them as comma-separated text? Because in read mode the tool always returns them as an array, even though internally the PDF only stores a single string. It's a practical approximation, not an exact byte-for-byte conversion.

What happens if I get a date format wrong? You get a 400 error. Dates need to be in ISO 8601 format, like 2026-08-06T10:00:00Z.

How much does it cost to read or change the metadata? 1 credit per call, whether you're reading or writing.

Try these tools

You'll find the full instructions for PDF metadata, covering both modes and every available field. And if you don't have an account yet, you can create one for free.