How to Compress a PDF and Reduce Its File Size

What makes a PDF big

Almost always: pictures. A scan or a brochure full of photos can be tens of megabytes, because each picture is stored in high resolution and high quality. Text, on the other hand, is cheap. A hundred pages of text are often well below one megabyte.

There is also some hidden weight: the same font or the same picture stored several times, small preview pictures of every page (thumbnails), and data that the program which made the file left behind. This is why two PDFs of the same document can differ a lot in size.

Compress a PDF step by step

1. Open the Compress PDF tool and add one or more PDFs.
2. Choose a compression level. Low keeps the most quality, Maximum gives the smallest file. Medium is the default and is a good place to start.
3. If a form tells you the limit, type it into Target size (MB). The tool then tries stronger and stronger settings until the file fits, and stops at the first one that does, so you do not lose more quality than needed.
4. Press Compress PDF. One file is saved as a PDF, several files come as a ZIP. The size before and after is shown next to each file.

Compressing runs in your browser, so the file is not uploaded.

What the levels really change

The level sets two things for the pictures in the file: the JPEG quality and the highest resolution (dpi) they are allowed to keep. Low is quality 75 at 200 dpi, Medium is 50 at 150 dpi, High is 35 at 110 dpi and Maximum is 20 at 72 dpi. For reading on a screen, Medium is usually fine. For something that will be printed, stay with Low. 72 dpi is only good for a screen and can look soft or blocky on paper.

Text and drawings are not touched. Only the pictures get lighter, so a photo can look a little softer, but the letters stay sharp.

In one of our own tests, a 3.7 MB test PDF with two pages of large photos came down to about 210 KB at the default level. That was a very favourable case. A PDF that is mostly text has little to gain, and a PDF whose pictures are already small will hardly change.

Lossless optimize

If you do not want the pictures touched at all, choose Lossless optimize in the Compression Algorithm box. The pictures stay exactly as they are. What changes is everything around them: identical pictures and fonts are stored once, unused objects are removed and the data is packed more tightly. You will save less, but nothing about how the pages look changes.

Rasterize: only for scans

The Rasterize option turns every page into a picture. For a scan that is already a picture, this can shrink the file a lot. For a normal PDF it would turn the text into a picture too, so you could no longer select or search it. For that reason the tool refuses to rasterize a PDF that still contains a text layer, and tells you which page it found text on.

If the file will not shrink

• The result is never bigger than what you put in. When nothing smaller can be made, you get your original file back, and the tool says so.
• If the target size cannot be reached, you are told the smallest size it did reach. Then you have a choice: accept it, split the file in two, or remove pages you do not need.
• A file with a password cannot be compressed as it is. Unlock it first with the Lock & Unlock PDF tool, then compress the unlocked copy.
• Compressing needs memory on your own device. For a very large PDF, split it first and compress the parts.
• If the PDF must stay PDF/A, tick Preserve PDF/A. It keeps the metadata and colour information that PDF/A needs, and turns off the options that would break it.

Check before you send

Always open the compressed file and look at the pages with the most pictures. If they look worse than you can accept, go back one level (Medium to Low) and try again. Keep the original until the compressed file has been accepted.