How to make a PDF smaller, tested on real files
Why PDFs get heavy and which fixes really shrink them. We measured each method on the same files, including two popular tricks that make PDFs bigger.
Start with what is inside the file
A PDF is a container. Text and fonts are tiny: a page of plain text is usually a few kilobytes. What makes a PDF heavy is almost always pictures: photos, scans and screenshots placed on the pages. A scanned contract is nothing but pictures of paper, which is why a 10-page scan can weigh more than a 300-page novel.
So the question “how do I make this PDF smaller” is really “how do I make the pictures inside it smaller”. Every method below is a variation on that.
Method 1: shrink the photos before you build the PDF
If you are making the PDF yourself, from phone photos of receipts, forms or homework, this is the method that works best.
We took 9 sample photos (1,133 KB in total) and turned them into an A4 PDF with Image to PDF. The PDF came out at 1,137 KB, because the photos are placed inside without being re-encoded. Then we ran the same photos through Compress Image at quality 75 first. The PDF built from those came out at 554 KB, about half. Open one of the compressed photos and zoom in to check that any small print is still legible before you build the PDF.
A modern phone photo is much larger than our samples, often 3 to 8 MB, so the savings on real phone pictures are usually bigger. Setting Max size to 1,920 px in the compressor helps most: that is plenty for a page read on screen or printed on A4.
Method 2: send only the pages that are needed
The simplest way to make a file smaller is to leave things out. If someone asked for your ID and one payslip, they don’t need the whole 30-page bundle.
With Split PDF we pulled 10 pages out of a 100-page test file of 1,260 KB. The extract was 743 KB. It isn’t exactly a tenth, because pages can share images and fonts that have to come along, but it is a big cut for one minute of work. Organize PDF Pages does the same job when you want to remove a few blank or duplicate pages from the middle of a scan.
What doesn’t work: merging, and turning pages into images
Two tricks come up again and again online, and both disappointed in our tests.
Merging doesn’t compress. We combined 5 files totalling 5.6 MB with Merge PDF; the result was 5.6 MB. A good merge copies pages exactly, so the total is the sum of the parts.
Converting pages to JPG and back often makes things worse. We rendered a 1,148 KB, 10-page document with text and photos at 150 DPI using PDF to JPG/PNG. The ten JPGs alone came to 3,311 KB, before even putting them back into a PDF, and the text was no longer selectable or searchable. This route only helps for huge scans made at very high resolution, and even then it costs you searchable text.
A quick checklist
- Making a PDF from photos? Compress the photos first, then build the PDF.
- Sending part of a document? Extract only the pages that were asked for.
- Combining files? Expect the result to be as big as all the inputs together, and shrink the inputs instead.
- Got a heavy scan from a copier? Rescan at 150 to 200 DPI in greyscale if you can; it is the most effective fix of all.
All the tools mentioned here run in your browser, so contracts and payslips aren’t uploaded anywhere while you work on them.