PDF Metadata Cleaner

METADATA PDF PRIVACY
Share:
By Checked against sources by , Founder & Architect, RECATOOLS Published Updated Last reviewed

View, edit, or strip every metadata field from a PDF — author, creation date, producer, and more. 100% in your browser.

RT-IMG-016 · Image & File
Your browser is too old for this tool’s interactive features. Please update to Safari 13.1, Chrome or Edge 80, Firefox 74 or newer.

PDF Metadata Cleaner

No PDF loaded yet.
💡 What gets stripped: the document's /Info dictionary — Title, Author, Subject, Keywords, Creator, Producer, Creation date, Modification date. What does not: the XMP metadata stream (the XML block Adobe apps write and PDF 2.0 prefers) is left exactly as it was — pdf-lib 1.17.1 has no XMP support — so a Title or Author stored there survives. Check the saved file with exiftool or pdfinfo before relying on it. The page content itself is untouched.
Drop a PDF to begin.
📅 Research current as of 13 Sep 2026 · Sources: ISO 32000-1:2008 §14.3 / ISO 32000-2:2020 (Info dictionary, XMP), ISO 16684-1:2019, pdf-lib 1.17.1
Standards, published formulas, and reference data are revised over time — check against a current authoritative source before relying on these figures.
🔒 PDFs stay on your device. Metadata editing happens entirely in your browser using pdf-lib 1.17.1, served from this domain. Nothing is uploaded — verify in DevTools → Network.
Advertisement
After results · AD-W1 Responsive · Post-tool — peak engagement

How to clean PDF metadata

Add your PDF

Drag your file onto the dropzone or click to choose. The tool reads the metadata and shows every standard field — Title, Author, Subject, Keywords, Creator, Producer, and both dates.

Review what's in there

Many PDFs carry surprising fingerprints — the original Word document's author name, the precise creation timestamp, the version of Acrobat Pro used to export it. Confidential? Embarrassing? Either way: now you can see it.

Edit or strip

Type into any field to change it. Use Strip all metadata to clear every field at once — useful before sharing a PDF outside your organisation. Restore original undoes your edits if you change your mind.

Save the scrubbed PDF

Click Save scrubbed PDF. The downloaded file has the Info fields you specified (empty strings if you stripped; a cleared date is written as 1 January 1970). The page content — text, images, layout — is unchanged, although pdf-lib re-serialises the whole file, so the bytes and the size will differ from the original.

Advertisement
After how-to · AD-W2 Responsive

PDF metadata — the fingerprint you didn't know you left

Every PDF carries a metadata block alongside the visible page content. When you export a PDF from Microsoft Word, the document's author name (the Windows account that owns the Word installation) is copied into the PDF's Author field. When you scan a document on a multi-function printer, the device's serial number and firmware version often end up in the Producer field. When you fill in a form in Acrobat and save it, your installed Acrobat license name lands in the Creator field. None of this is visible when you open the PDF — but it's all there in the file, readable by any tool that knows where to look.

Why this matters more than people think

A redacted PDF whose Author field still names the person who drafted it is the classic metadata leak: the visible content was scrubbed, the Info dictionary was not. Adobe Acrobat Pro has a "Sanitize Document" command for exactly this reason. This tool covers the first half of that job — showing you the Info fields and letting you clear them — and is explicit about the half it does not cover.

PDF 2.0 (ISO 32000-2) deprecates every Info-dictionary entry except the two dates in favour of XMP — which is exactly the stream this tool cannot yet clear. Read the "What about XMP metadata?" answer before you trust a scrubbed file.

What this tool does — and what it doesn't

This tool edits or clears the standard PDF /Info dictionary — the eight text and date fields shown in the form (the ninth Info key, Trapped, is a prepress flag and is left alone). What it does NOT touch: the XMP metadata stream — pdf-lib 1.17.1 neither reads nor writes it, so a Title, Author or creator tool recorded there by Word, InDesign or Acrobat stays in the file — the page content itself (text, images, layout), embedded fonts (which can carry vendor signatures), JavaScript actions, form-field values, embedded files (attachments), or comments and annotations (each of which has its own author entry). For complete sanitisation including those layers, use Adobe Acrobat's Sanitize Document, or exiftool -all= file.pdf for the XMP stream, and verify with pdfinfo or exiftool afterwards.

Nine Info keys, a 1970 date, and the XMP stream this tool cannot reach

01

ISO 32000-1 (Table 317) defines nine keys in the document information dictionary: Title, Author, Subject, Keywords, Creator, Producer, CreationDate, ModDate and Trapped. This tool edits the eight text and date keys.

02

Adobe introduced XMP in 2001; it is now ISO 16684-1:2019, and PDF 2.0 (ISO 32000-2) deprecates every Info entry except CreationDate and ModDate in favour of it. A modern PDF often carries the same author in both places — and this tool clears only one of them.

03

The Producer field tells you the EXACT software that produced the PDF — including version numbers. "Microsoft Word 365 (16.0.17328.0)" tells an attacker more than you might want.

04

The Author field defaults to the OS account name on Windows. Many PDFs share a user's real name (or worse, internal username) without the user ever noticing.

05

Creation date precision is to the second, including timezone. That's enough to correlate document authorship with calendar events or surveillance logs.

06

Modification date updates every time you save a PDF, including from Preview on macOS or a browser's PDF viewer. Tools sometimes leak edit timestamps unintentionally.

07

PDF/A (ISO 19005) requires an XMP metadata stream and requires any Info entries present to agree with it. Clearing Info while XMP keeps the old values breaks that agreement — if you need PDF/A conformance, edit consistently rather than strip.

08

A cleared date is written as 1 January 1970 00:00 UTC (D:19700101000000Z) rather than removed — pdf-lib offers no way to delete an Info key, so the Unix epoch is the unambiguous "cleared" marker. A reader will show that date, not a blank.

09

Annotations, form fields, and digital signatures carry their own metadata — including the signer's identity. This tool doesn't touch those, only the document-level metadata.

10

Producer strings fingerprint the exporter: "Microsoft® Word for Microsoft 365" or "Skia/PDF m118" (Chrome's print-to-PDF) each name a program and a version. Leave the field blank and the output carries an empty Producer — this tool does not write its own name into it.

FAQ

  • No. The pdf-lib library runs entirely in your browser. Your PDF is read into memory, metadata is rewritten in memory, and saved back to your downloads. Open DevTools → Network and watch — there's no outbound traffic.

  • No. The /Info dictionary is informational — PDF readers don't depend on it to display the document. Stripping every field is safe for any PDF that doesn't claim PDF/A compliance.

  • It is not touched. pdf-lib 1.17.1 has no XMP support: the metadata stream referenced from the document catalogue is copied through unchanged on save, so a dc:creator, dc:title or xmp:CreatorTool written there by Word, InDesign or Acrobat survives your edits — and PDF 2.0 readers prefer XMP over the Info dictionary. Until this tool can rewrite XMP, treat its output as "Info cleaned, XMP unknown" and check with exiftool file.pdf or pdfinfo -meta file.pdf; exiftool -all= file.pdf clears both.

  • No. Annotations (comments, highlights, sticky notes) and form field properties have their own author metadata that this tool does NOT touch. For complete sanitisation including those layers, use Adobe Acrobat's "Sanitize Document" feature.

  • No. PDFs can contain embedded file attachments (other PDFs, spreadsheets, images, even executables). Those carry their own metadata that this tool doesn't reach. The /Info dictionary is document-level only.

  • YYYY-MM-DD or YYYY-MM-DD HH:MM (both work; an ISO 8601 string also parses). An empty field does not remove the key — pdf-lib cannot delete Info entries — so it is written as the Unix epoch, 1 January 1970 00:00 UTC, which a reader will display as that date. Dates are saved in the PDF date format (D:YYYYMMDDHHmmSS+TZ).

  • No. pdf-lib 1.17.1 does not support encrypted PDFs, including "restriction-only" (owner-password) files whose strings and streams are still encrypted. The tool refuses them at load (it used to accept them without an error), because pdf-lib's own documentation says the result of modifying such a document "may fail, or have unexpected results" — the rewritten Info strings would be written unencrypted into a file that declares encryption, and readers reject that. Remove the encryption first (Acrobat, macOS Preview export, or qpdf --decrypt).

  • Yes — type into any field. The PDF spec also allows arbitrary custom keys in the /Info dictionary, but those aren't covered by this tool. For custom keys you need Adobe Acrobat's Document Properties dialog.

  • Usually a little, sometimes noticeably: pdf-lib re-serialises the entire file on save — objects are renumbered and cross-reference tables rebuilt — so the size can move by more than the few hundred bytes of metadata you changed, in either direction. The page content is identical.

  • Yes on iOS Safari and Chrome on Android. The form reflows to a single column on phone-width screens. All editing happens locally — same privacy story as desktop.

Method & sources

How it computes

Loads the PDF with pdf-lib 1.17.1 in the browser, reads the eight text and date entries of the document information dictionary (ISO 32000 §14.3.3: Title, Author, Subject, Keywords, Creator, Producer, CreationDate, ModDate) into an editable form, and on save writes the edited values back with setTitle()…setModificationDate() and re-serialises the file as {name}-scrubbed.pdf. Clearing a text field writes an empty string;

What this tool implements

  • Edits the Info dictionary only: eight of the nine keys ISO 32000-1 Table 317 defines (Trapped is left alone), read and written through pdf-lib's typed accessors.
  • Cleared dates are written as 1 January 1970 00:00 UTC rather than removed — stated on the page, because a reader will display that date, not a blank.
  • XMP is out of scope and declared so: ISO 32000-2 deprecates every Info entry except CreationDate and ModDate in favour of XMP (ISO 16684-1), so a modern PDF may carry the same author in a stream this tool does not touch; the page tells the reader how to check (exiftool, pdfinfo) and how to clear it.
  • PDF/A (ISO 19005) requires an XMP stream and requires Info entries to agree with it — the page warns that stripping Info alone breaks that agreement.
  • Encrypted input is not supported: the tool loads with ignoreEncryption:true, but pdf-lib's documentation says this does not decrypt and results 'may fail, or have unexpected results'; the page tells the reader to decrypt first.
  • Annotation, form-field and signature author entries live in their own dictionaries, not in Info, and are untouched — stated in the FAQ.

Sources

What can make this go out of date

  • none at runtime — a pdf-lib release adding XMP support or Info-key removal would let the tool clear what it currently cannot; the page's limits are tied to v1.17.1.

Abridged — the full review record for this tool runs longer than the list above.

Advertisement
Pre-footer · AD-W3 728 × 90