How to compress a PDF
- Choose your PDF. Drop it onto the box or tap to pick it. Compression starts right away with the Balanced level.
- Read the result. The bars show the size before and after, and the line under them says what changed: how many images were re-encoded, and why any were left alone.
- Check a page. The compare view shows the same page from the original and the compressed file side by side. Tap a page to see it at full size, and step through the pages with the arrows.
- Adjust and download. Switch to Light or Strong, or turn photos black and white; the PDF is compressed again in a moment. Then download it.
How the compression works
In most large PDFs the weight is in the pictures: scanned pages, photos and screenshots saved at full camera resolution and high quality. The compressor finds every image in the file and, for each photo, works out the largest size that is still useful on the page: 220, 150 or 100 pixels per inch of the largest page, depending on the level. Bigger images are scaled down to that, then saved as JPEG at a quality that suits the level. An image is only replaced when the new version is clearly smaller, so nothing gets worse for no gain.
Everything that isn't a photo is left exactly as it is. Text, fonts, lines and vector drawings keep their full sharpness, and images that look like charts, diagrams or screenshots stay lossless, because JPEG would put fuzz around their edges. Black-and-white scans stored with CCITT or JBIG2 are already tiny and are kept too. On top of that, the tool compresses streams that were stored without compression, drops objects that nothing in the file uses any more (left over from earlier edits), and saves the file with compact object streams.
| Level | Photos at most | Good for |
|---|---|---|
| Light | 220 dpi, high quality | Photos you will print, brochures |
| Balanced | 150 dpi | Most documents: on screen and office printing |
| Strong | 100 dpi, lower quality | Email attachments and upload forms with a size limit |
When a PDF can't be made smaller
Some PDFs are already as small as they will get: files exported from Word or Google Docs without pictures, forms, and documents that were compressed before. Their size is text and fonts, which are already stored compressed. Instead of giving you a file that is the same size or bigger, the tool tells you so and doesn't offer the download. For a scan that only just misses a limit, Strong plus Turn photos black and white usually finds the last few hundred kilobytes, because a grey image needs far less data than a colour one.
Private by design
Your PDF is read and rewritten by your own browser on this page, with pdf-lib and the browser's built-in JPEG encoder; the page preview is drawn with pdf.js. Nothing is uploaded, so you can compress scans of passports, contracts and medical records without them leaving your device. A progress bar shows how far it has got; long scans take longer on a phone than on a computer.
Troubleshooting
- "Editing restrictions" or "needs a password": a restricted file can be compressed as an unlocked copy with one tap; for one that needs a password to open, print it to a new PDF in a PDF app first.
- Photos look blocky: go back to Balanced or Light. The compare view at full size shows the difference best on faces and fine print inside photos.
- Still too big for an upload form: split the job. Turn pages into images with PDF to JPG and shrink them with the image compressor, which can aim at an exact size like 200 KB.
- Several files to send as one: combine them with Merge PDF first, then compress the merged file once.