How to reduce PDF file size, and what actually works

27 August 2026FreePanda Team

A PDF that will not fit an upload limit is one of those problems where the obvious fix — running it through a compressor — sometimes takes 60% off and sometimes achieves nothing at all. The difference is entirely down to what is inside the file.

First, find out what your PDF is made of

There are broadly two kinds.

Text and vector PDFs come out of word processors, spreadsheets and design tools. The text is stored as characters plus font data, and the graphics as instructions. These files are small to begin with and their remaining bulk is structural: embedded fonts, object overhead, metadata.

Image PDFs come from scanners and phone cameras. Every page is a photograph. A ten-page scan can easily be 20 MB, and effectively all of it is image data.

If you can select the text with your cursor, it is the first kind. If you cannot, it is the second — and that determines which method will work.

Method 1: Rewrite the structure

This repacks the document's internal objects into compressed streams and strips metadata. It is completely lossless — nothing is decoded, downsampled or re-encoded, so text stays selectable and vectors stay sharp at any zoom.

On a text-heavy PDF this can be a substantial saving. On a scan it will barely register, because there is almost no structural overhead to remove relative to the image data.

This is what the PDF Compressor does, and it deliberately stops there rather than touching your images without asking.

Method 2: Compress the images before they go in

For an image PDF, this is the only method that matters. The images are the file.

The order that works:

Extract or re-export the images.
Run them through the Image Compressor, or the Resize Image to KB tool if you have a specific target.
Rebuild the PDF with the smaller images.

A 300 DPI scan viewed on a screen is carrying roughly four times the pixels it needs. Dropping to 150 DPI typically cuts the file by three quarters with no visible difference on screen — though it will look softer in print.

Method 3: Scan at a sensible setting in the first place

Most of the oversized PDFs people struggle with were created at settings nobody chose deliberately.

Scan in greyscale, not colour, for a black-and-white document. It removes two thirds of the data immediately.
150–200 DPI is enough for anything that will be read on a screen. 600 DPI is for archival reproduction.
Use your scanner's "document" mode rather than "photo" — it applies compression suited to text.
On a phone, use a scanning app rather than the camera. It flattens, crops and compresses appropriately, where the camera saves a full-resolution photograph.

Method 4: Remove what you do not need

Split out the pages you actually need. A 40-page report where the recipient wants pages 12–14 is best solved with the PDF Splitter, not a compressor.
Flatten form fields if the form is filled and will not be edited again.
Drop embedded attachments, which some PDFs carry invisibly.

Hitting a specific limit

Common ceilings: 25 MB for Gmail attachments, and many government and university portals ask for 2 MB or less — sometimes as little as 500 KB.

Work backwards. If a 10-page scan must fit 2 MB, that is 200 KB a page, which means roughly 150 DPI greyscale JPEG. Compress the images to that target first, then assemble. Trying to squeeze the finished PDF is working on the wrong end of the problem.

What to be wary of

Any tool promising a huge reduction on a scanned PDF is re-encoding your images, and that is a quality decision it is making on your behalf. Sometimes that is exactly what you want. But it is irreversible, and if the document is a signed contract or an identity proof, softened text can matter.

Be wary too of where the file goes. A PDF you need to shrink is very often a bank statement, a contract or an ID document — precisely the sort of thing that should not be uploaded to an unknown server to be processed. Everything on FreePanda runs in your own browser.