Export Annotations to CSV / JSON - Free Online Browser Tool
Dump every comment, highlight and stamp in a PDF to a spreadsheet or a data file.
The comment layer of a PDF is where a review actually lives, and it is the one part no ordinary text extraction will show you. Twenty reviewer notes, eleven highlights, a few stamps - and a question that needs answering: who said what, on which page, and when. Scrolling through is how you read three notes. Answering "did we miss anything from round two" across two hundred pages is not a scrolling job. This puts the whole layer in a table. Page, type, author, date, position and the comment text, as CSV that opens in any spreadsheet or JSON if you are feeding it into something else. Three decisions are worth stating, because each one deliberately loses or keeps something. Links are excluded: a link is a pointer to somewhere else, not a note a person wrote, and a tracker listing four hundred of them tells you nothing about the review. An annotation with no written comment is still included - it existed, it is on the page, and dropping it would make the count a lie. And an annotation with no date stays undated in the export rather than being given one, because "never dated" and "dated in 1970" are different facts and only one of them is true. A half-hour timezone offset is kept rather than flattened to UTC, so a reviewer in Chennai and one in London do not end up with the same timestamp. The summary counts what came out and groups it by author, which is usually the first thing anyone wants to know. One more thing this does on the way: a comment that begins with an equals sign, a plus or a minus is prefixed so a spreadsheet treats it as text. Opening an exported file and having Excel try to evaluate a note as a formula is a small thing that has burned people. A PDF with nothing marked on it produces an error saying so, rather than an empty file that looks like the tool failed quietly. If you want the text layer instead of the markup, Compare Extracted Text is the one for differences between versions; this is strictly for what reviewers wrote. Everything runs in this tab.
This export annotations to csv / json utility runs completely inside your web browser. Your document is opened locally and processed on your own device using JavaScript and WebAssembly, so your file is never uploaded to any server. There is no signup, no watermark and no limit on how many documents you can handle.
- 100% client-side - files never leave your device
- Instant processing, no upload wait
- Works on desktop, tablet and mobile
- Free forever, no account required
How to Use the Export Annotations to CSV / JSON Tool
Four steps, and the whole thing runs locally on your own device.
-
Add the marked-up PDF
The one carrying the review. Reading annotations needs no rendering, so this stays quick even on a long document.
-
Choose the format
CSV opens in any spreadsheet and is the right choice for reading. JSON keeps the structure intact if you are passing it to something else.
-
Choose what to include
Everything except links, or only the annotations that carry a written comment. The second is the quicker read when you are chasing one reviewer.
-
Save
One ZIP: the table, plus a plain-text report giving the total and a breakdown by author.
Frequently Asked Questions
Why are links left out?
A link points somewhere else; it is not something a reviewer wrote. A tracker listing every URL in a document tells you nothing about the review that happened on it, and burying the actual comments among them makes the export harder to use, not easier.
What happens to a highlight with no written comment?
It is still included, with an empty comment cell. It is on the page, so it counts - dropping it would make the number of annotations in the file disagree with the number in the export, which is exactly the kind of quiet inaccuracy that gets a report questioned.
Do the dates come out in my timezone?
They come out as written in the file, with the offset preserved rather than converted. An annotation saved at midday in a half-hour timezone keeps that offset instead of being flattened to UTC, so two reviewers in different zones do not end up looking as though they commented at the same moment.
Can I open the result in Excel without anything going wrong?
Yes. Comments that begin with an equals sign, a plus or a minus are prefixed so the spreadsheet reads them as text rather than trying to calculate them, and commas, quotes and line breaks inside a comment are quoted properly rather than splitting the row.
It says there are no annotations - is that right?
Usually yes, and it is worth checking the file you picked. Comments made in a reader that saves them privately, or on a scan with no markup at all, will not be there. If you were expecting notes from a specific review, they may have been flattened into the page, in which case no annotation tool will find them.