Compress PDF

Compress a PDF

Compression is not one operation, so this does not pretend it is. A lossless rewrite keeps every bit of text, link and form field. Flattening pages to images shrinks a scan enormously and turns its text into pixels. Automatic runs what makes sense for your file and gives you the smallest result — or your original back, if nothing beat it.

Drop a file here, or click to choose

One PDF

Mode

Runs a lossless rewrite, and additionally tries flattening the pages to images when the document looks like a scan. You get whichever result is smallest — and if neither beats your original, you get the original back untouched rather than a bigger file.

How to compress a pdf

  1. 1

    Drop in your PDF

    The file is read straight into the page. Nothing is uploaded, so there is no queue and no size cap beyond your device's memory.

  2. 2

    Leave it on Automatic

    It rewrites the file losslessly, and additionally tries flattening when the document looks like a scan. You can force either method instead if you know what your file is.

  3. 3

    Compress

    The result reports the size, the percentage saved, and exactly which method won — so you always know whether the text survived.

  4. 4

    Download

    If nothing beat your original, that is what you get back, unchanged, with an explanation rather than a worse file.

It will never make your file bigger

Every candidate is measured against the file you supplied, and the original wins ties. This matters more than it sounds: naively re-encoding pages as images inflates an ordinary text document by roughly 150 times per page, which is how a lot of “compressors” quietly hand back something worse.

Where a PDF's weight actually lives

Text is cheap — a page of it costs about a kilobyte. Almost all of a large PDF is embedded images: phone photos at full sensor resolution, or scans at 600 DPI. That is why a scanned invoice bundle compresses by 80% while a text report compresses by nothing at all.

You are told which trade you got

Lossless means object streams are rebuilt and page content copied verbatim. Flattening means each page became a JPEG, so text is no longer selectable and form fields and links are gone. Both are legitimate results; not saying which one you received is not.

It never leaves your device

Rendering and re-encoding both happen in the browser. Your document is not uploaded, queued, or held on a server for a couple of hours — there is no server in the path at all.

Questions

It depends entirely on what is in it. A scanned document typically drops by 70–90%, because full-resolution page images are replaced by page-sized JPEGs. A text-only PDF usually cannot be compressed at all — it is already about a kilobyte a page, and there is nothing left to remove.

Quill

A calm, fast PDF workspace. Free to use, private by default.

Product

EditorAll PDF tools

Templates

Quill AI

Mobile

Tools

Convert

Organise

Company

About

Privacy

Terms

Contact

© 2026 Quill. All rights reserved.
PrivacyTermsStatus