MinoPDFSanitize PDF
Remove metadata, comments, highlights, links and form values from a PDF in one pass, then download a new file with a report of what is still there. Runs in your browser.
Runs in your browser
Files are never uploaded — they stay on your device.
No page limit
Any document size, no per-page caps or quotas.
No watermarks
Clean output with no branding added.
No sign-up
Just open the tool and use it — no email, no account.
How to verify
Verify in 3 steps
No promises to take on faith — check it yourself in about 30 seconds.
- 1Open DevTools → Network tab
- 2Add a file and run the tool
- 3Confirm: no request carries your file
Once the page has loaded you can also unplug the network — every tool keeps working.
Full method & limitsKey Features
Six Categories in One Pass
Metadata and XMP, document ID, comments, highlights and underlines, links, and form values — pick what to clean and run once.
Reuses the Tools You Already Trust
Each category is handled by the same kernel as its dedicated tool page, so the result matches what you would get by running them one by one.
A Report of What Is Still There
The download comes with a report listing what was removed, what remains, and what needs another tool. It never says done when something is left.
Your Original Is Untouched
Nothing is written back to the file you dropped in — the download is always a new copy.
Honest About What It Cannot Do
Body text, embedded JavaScript, attachments and unembedded fonts are not removed here. They are listed as still there instead of being quietly ignored.
No Upload
Cleaning runs in your browser. There is no endpoint that receives the file, and the report contains counts only.
How to Use
Keep the original
The download is a new copy, but keep the original somewhere safe — cleaning is easy to redo and hard to undo by hand.
Add the PDF
Drop the file in. There is nothing to configure before the first run beyond picking the categories.
Pick the categories
Metadata, comments, highlights, links and form values are separate switches. Turn off anything you want to keep.
Clean and read the report
Download the new file and check what was removed, what remains and what needs the redaction or flatten tool.
Verify if the file is sensitive
Drop the result into the inspect tool and scan it again — before and after should differ exactly where you expect.
Cleaning a Reviewed Draft
A paper draft has been commented on by two colleagues and is about to go to an outside reviewer.
manuscript-draft.pdf (24 pages)
The file carries author metadata, an XMP stream, 14 comments with reviewer names, 11 highlights, 3 external links and 2 filled form fields.
manuscript-draft-cleaned.pdf + clean-report.json
Metadata fields, comments, highlights, links and form values are gone; the report states that one attachment and one embedded script are still there, because no tool on this site removes them yet.
How to Clean a PDF Before Sharing It
How to Clean a PDF Before Sharing It
A PDF carries more than the pages you see. Document properties, reviewer comments, links, form values and scripts travel with the file and survive copying, renaming and exporting. Cleaning a PDF means removing that residue in one pass and then checking what is left.
Quick Answer
Add the PDF, tick the categories you want gone — metadata and XMP, comments, highlights, links, form values — then clean and download. The download comes with a report listing what was removed, what remains and what needs another tool.
What Cleaning Removes
- Document properties and XMP: author, title, subject, creator, producer, dates, plus the second metadata copy that most viewers actually read.
- Document ID: the trailer identifier that can still link copies of the same file together.
- Comments and highlights: reviewer names, dates and the marks they left.
- Links: external addresses, mail links and internal jumps.
- Form values: what was typed into form fields, without flattening the form.
Cleaning Is Not the Same as Redaction
Cleaning removes hidden structural residue. Redaction removes visible content: a name in a sentence, an account number in a table, a signature at the bottom of a page. Adobe keeps these as two separate features for the same reason — sanitize for hidden content, redact for visible content. If the sensitive information is in the body of the document, cleaning will not remove it.
Why Metadata Alone Is Not Enough
Most tools stop at the Info dictionary. The XMP stream is a second copy that readers prefer, the document ID still links copies together, and comments live in their own objects. Cleaning only one of them leaves the rest of the trail intact.
What This Tool Cannot Remove
Body text, embedded JavaScript, attachments, unembedded fonts and pages without a text layer are outside this page. They are listed in the report under what is still there. Body text goes to the redaction tool; scripts and attachments currently need manual handling.
How to Verify the Result
Read the report, then drop the cleaned file into the inspect tool and scan it again. The before and after counts should differ exactly where you expected. For anything going to an AI service, run the same check first and remove what the model does not need.
Keep the Original File
Nothing is written back to the file you added, so the download is always a copy. Keep the original somewhere safe: cleaning is easy to redo and awkward to undo by hand.