How to Compress a PDF Without Wrecking the Quality
Published 22 August 2026 · 6 min read
Compression is usually approached as one button pressed repeatedly until either the file fits or it becomes unreadable. It works better if you first find out what is actually large, because the right fix for a scanned document is the wrong fix for a report full of charts.
What is taking up the space
Text is tiny. A hundred pages of prose is a few hundred kilobytes. If your PDF is large, it is almost always one of these:
Images. By far the most common cause. A single phone photo can outweigh a hundred pages of text, and scanned documents are entirely images.
Embedded fonts. A full font family adds hundreds of kilobytes. Usually only a problem when many fonts are embedded unsubsetted.
Form fields and annotations. Stored as live objects with their own structure.
Revision history. PDFs support incremental saving, where each edit appends to the file and the old version stays. A heavily edited document can carry several copies of itself.
Vector artwork. Charts and maps with tens of thousands of paths can be surprisingly heavy.
A quick diagnostic: extract the images from the document. If they account for most of the file size, it is an image problem — which is good news, because image problems have the best fixes.
The order that preserves quality
Work through these in sequence and stop when the file is small enough. Each step costs less quality than the one after it.
Remove what you do not need.Delete pages you are not sending. The cheapest byte is the one that is not there.
Flatten forms and annotations.Flattening converts live objects into page content, often shrinking the file and removing recoverable data at the same time.
Drop colour if it carries no information.Greyscale conversion is the most underrated step — for a black-ink scan it can cut the size dramatically and looks identical when printed.
Compress the images. Now run PDF compression. This is where quality is actually traded away, which is why it comes after the free wins.
Only then, reduce resolution. If it still will not fit, the images are simply at a higher resolution than needed. 300 DPI is right for print; 150 is fine for screen reading.
Most people start at step four and never discover that steps one to three would have been enough on their own — without touching quality at all.
Where quality actually breaks
Over-compression does not degrade a page evenly. It fails first and worst on exactly the parts that matter:
Small text in scans becomes fuzzy and stops being OCR-readable long before it looks bad at full-page zoom.
Fine lines in tables, signatures and technical drawings break up.
Sharp edges — text against a flat background — get haloes, because JPEG-style compression is built for photographs, not for letterforms.
Gradients band into visible steps.
So do not judge the result by scrolling at page-fit zoom. Zoom to 200% on the smallest text on the busiest page. If that is legible, the document is fine. If it is not, back off one quality level.
When compressing is the wrong answer
The file is huge because it should be several files. A 300-page manual emailed to someone who needs chapter 4 should be split, not compressed.
It is for print. Printers need the resolution. Compressing a document destined for a press is solving the wrong problem.
It is an archival or legal copy. Compression is lossy; the discarded detail does not come back. Keep the original.
The recipient just needs the text.Extract the text and send that — a fraction of the size and far easier to work with.
A note on where this happens
Compression is normally the last step before sending a document to someone — which means running it through an upload-based service puts the finished, complete document on a third party's servers at the exact moment it is most sensitive. Compressing locally avoids that, and you can verify nothing is transmitted while it runs.
Try it yourself
Compress PDF — runs in your browser, nothing is uploaded.
Frequently asked questions
Why is my PDF so large?
Almost always images. Text is very small — a hundred pages of prose is a few hundred kilobytes — while a single photograph or scanned page can be several megabytes. Embedded fonts, form fields and retained revision history are less common causes.
How can I compress a PDF without losing quality?
Do the free steps first: delete pages you do not need, flatten forms and annotations, and convert to greyscale if colour carries no information. Those often suffice. Only then compress the images, which is the step that actually trades quality away.
How do I know if I compressed too much?
Zoom to 200% on the smallest text on the busiest page, rather than judging at page-fit zoom. Small text and fine lines degrade first. If the small print is not comfortably legible, step back one quality level.
Does converting a PDF to greyscale really reduce the size?
Often substantially, because colour images store roughly three times the channel data of greyscale ones. For a document that is black ink on white paper it is close to a free saving, and it prints identically on a mono printer.