Reports and research
Your paid analysis circulates as PDFs. Protection discourages automated reuse without getting in your readers’ way.
Anti-scraping for PDF documentsYour PDFs stay readable.
Your PDFs stay readable.
Bots come away with nothing.
Two levels to choose from. The first slips invisible text into your documents, blended into the exact color of the background: readers see nothing, scrapers pull out scrambled content. The second turns every page into an image — leaving no text to extract at all.
- 100%
- of the rendering preserved
- 0 px
- of visible difference
- 2
- protection levels
rapport-t3.pdf
Quarterly Report — Q3 2026
Revenue is up 18% over the quarter,
carried by subscription growth.
Gross margin reaches 61%, four points
higher than in Q2.
How it works
Two levels, depending on what you are protecting
The first leaves the document untouched and scrambles what comes out of it. The second removes the text layer: nothing left to extract.
Invisible text
No watermark, no flattened image, no loss of quality.
- 1
Reading the background
The color fills on each page are read straight from the document: a solid background, a header band, a colored column. Every area is mapped with its coordinates.
- 2
Character-by-character blending
Each letter takes the exact color of the area it covers, not an average shade. Kerning is preserved, so the rendering stays identical to the original.
- 3
Scrambled extraction
The injected text is laced with homoglyphs and zero-width characters, and written out of order. Readers see nothing; extractors get back an unusable string.
Pages as images
The strongest protection, at the cost of text selection.
- 1
Rendering every page
Each page is rendered to an image at 150 dpi, a resolution that stays sharp on screen and in everyday printing.
- 2
Rebuilding the PDF
The images are reassembled into a fresh PDF at the exact dimensions of the original. The document opens, scrolls and prints just as before.
- 3
No text left
No text layer remains: extractors come back with an empty string. Text selection is gone and the file grows heavier — that is the price of this level.
Use cases
Built for documents that carry value
Contracts and proposals
Quotes, sales proposals, terms: flag the protection explicitly and make reuse traceable.
Educational material
Course materials and textbooks stay perfectly readable, but hold up against bulk harvesting.
Licensed publications
White papers and research notes published online, protected before they go out.
FAQ
Frequently asked questions
What is the difference between the two protection levels?
Invisible text leaves the document intact: its text stays selectable, but what an extractor pulls out is scrambled. Converting pages to images removes the text layer entirely — there is nothing left to extract, at the cost of text selection and a heavier file. The two do not stack: converting to images also erases the injected text.
Does the document still look identical?
Yes. In invisible-text mode, each character takes the color of the background it covers, read from the document itself: on screen and in print, nothing shows. In image mode, pages are rendered at 150 dpi, which stays sharp on screen and in everyday printing.
Does this really stop an AI from reading my document?
No, and that deserves a straight answer. These protections break the pipelines that read a PDF’s text layer, which is still the most common route. OCR or a model that can analyze an image will get past both modes. No transformation of a document a human can read will stop a machine from reading it the same way.
Is the text injected on every page?
On every page with room for it. Injected lines are placed where the page is clear, never on top of a heading or a paragraph: a letter laid over existing text would strike through it, since it takes the color of the background rather than of the letters it covers. A page filled top to bottom therefore gets no injection — better an unmarked page than a damaged document.
Is the injected text detectable?
A text extractor can read it — that is the point. Anyone who hits "select all" and copies would see the text appear. The protection targets automated analysis, not absolute concealment.
Does this break document accessibility?
This is the trade-off to know about, and it is real in both modes. A screen reader reads the text layer: in invisible-text mode it will also read out the injected content; in image mode it finds no text at all. For a document intended for visually impaired readers, we recommend enabling neither protection.
Which formats are accepted?
PDFs up to 10 MB. Invisible-text mode assumes a text layer: a fully scanned PDF has none, and is not readable by a standard extractor anyway. Image mode works on any PDF, but is capped at 60 pages — each rendered page weighs roughly 150 KB, so a longer document would produce an unwieldy file.
Are my files kept?
Processing happens on the fly and the file is deleted from our servers once the protected PDF is returned. No copy is archived.
Protect your first document
Drop in a PDF, pick your protection level, and get back a file that looks identical — but gives automated extraction nothing usable.
Drop a PDF here, or click to browse
PDF only — 10 MB maximum