How to compress a PDF
The short version: drop it into a compressor that recompresses the images and leaves the text alone. The longer version explains why that works, and why some PDFs refuse to budge.
What's taking up the space
A PDF is a container of objects. Text is stored as short drawing instructions ("put these glyphs at x, y") compressed with the same DEFLATE algorithm ZIP uses, so a page of text typically costs a few kilobytes. Fonts are embedded, but usually only the characters actually used. Images are stored as whole pictures, often at full camera or scanner resolution. That's where the megabytes live.
Change the sliders to see how the balance shifts between a typed report and a scanned one, and what each compression method can touch.
- Text and drawing instructions
- Embedded fonts
- Images and scans
Images carry almost all the weight, so re-encoding them does the work while text stays sharp, selectable and searchable.
Illustrative sizes based on typical files (about 4 KB of text per page, 450 KB per full-quality photo, 0.9 MB per 300 dpi scanned page). Your real file will differ; the tool above reports the actual result.
Step by step
- Try image recompression first. In the PDF compressor, keep "Keep text" and "Balanced". Photos are resized to a sensible maximum and saved as JPEG at a moderate quality.
- Use grayscale for black-and-white documents. A colour scan stores three channels; grayscale stores one.
- Only flatten if you must. If a portal demands 100 KB or 200 KB and the document is a scan, flattening at 72–110 dpi gets there, at the cost of selectable text.
- Fix it at the source next time. Scan at 150 dpi grayscale for forms, and compress photos before inserting them into Word or Slides.
Why a text-only PDF won't shrink
If your PDF came from Word or Google Docs and has no photos, it's probably already a few hundred kilobytes at most, and most of that is fonts. Every compressor will struggle, because there's nothing redundant left to remove. The only big lever is flattening, which trades a crisp, searchable document for a blurry picture of one. For a text PDF over 1 MB, look for a hidden culprit: a large logo image repeated on each page, or a high-resolution background.
A note on quality settings
JPEG quality isn't a percentage of the original. It's a dial for how coarsely the encoder rounds away fine detail in 8×8 pixel blocks. Most people can't see a difference between 100 and 80 on screen, while the file often halves. Below about 50, blocks and halos around text in scans start to show, which is why our Strong preset stops there and the target-size ladder only goes lower as a last resort.
Questions people ask
How do I make a PDF smaller for free?
Use a compressor that re-encodes the images inside it, such as the free tool on squeezefiles.com/compress-pdf, which works in your browser. On a Mac, Preview's Export > Quartz Filter > Reduce File Size also works, though it is often very aggressive.
Why is my PDF so big?
Almost always images: phone photos pasted into a document, or pages scanned at 300–600 dpi in colour. Embedded full fonts and duplicated images from careless exports add more.
Does compressing a PDF reduce quality?
Only the images lose quality, and at moderate settings the loss is hard to see on screen. Text and vector graphics are stored as instructions, not pixels, so they stay perfectly sharp unless you flatten the pages.
How do I compress a PDF in Word or Google Docs?
In Word, use File > Compress Pictures (Windows) or Reduce File Size (Mac) before saving as PDF, choosing 150 ppi. Google Docs has no setting; export, then run the PDF through a compressor.