"Compress without losing quality" is partly a contradiction — every lossy compressor trades some fidelity for size. The real goal is to remove waste you cannot see while leaving the parts you can. Understanding what actually makes a PDF large helps you compress smartly instead of blindly.
What makes a PDF big
In most documents, text and vector graphics are tiny; images dominate the file size. A phone photo dropped into a PDF can be 4000 pixels wide even though it only ever displays at letter width. That is pure waste. The two levers that matter are:
- Resolution (DPI). Downsampling a 600-DPI scan to 150 DPI for on-screen use can cut size dramatically with no visible change.
- JPEG quality. Re-encoding embedded images at a sensible quality level removes data the eye barely registers.
How PDFelly compresses safely
The Compress tool inspects each embedded image, downsamples ones that are larger than they need to be, and re-encodes them — but only keeps the result if it is actually smaller. It deliberately skips images it cannot touch safely (CMYK, transparency masks, indexed colour) so it never corrupts them, and it verifies the finished file still parses before handing it back. If compression would make the file bigger, it returns your original unchanged.
Choosing a level
- Screen / email — aggressive downsampling to ~150 DPI; smallest files, ideal for sharing.
- Balanced — a middle ground for documents you may print occasionally.
- Print — light touch that preserves higher resolution.
When compression barely helps
If a PDF is mostly selectable text, it is already small and there is little to gain. If a "PDF" is really a stack of full-page scanned images, consider running OCR so the text becomes searchable, then compressing. For email-specific size targets, see our guide on shrinking PDFs for email.
Match the compression level to the destination
The single best habit is to compress for where the file is going. A PDF you will only ever read on a screen can be downsampled hard with no visible difference, while one destined for a professional printer needs a gentle touch to protect fine detail. Guessing high "to be safe" wastes the entire benefit; guessing low can ruin a print. Decide the destination first, then pick the level.
Why two compressors give different results
If one tool halves a file and another barely changes it, they are making different choices about which images to touch and how aggressively. A cautious compressor refuses to re-encode images it could damage and keeps a result only when it is genuinely smaller. A reckless one re-encodes everything and can introduce visible blocking or, worse, corrupt unusual images. PDFelly errs toward caution and verifies the finished file still opens before handing it back.
Compressing scans versus born-digital PDFs
A "born-digital" PDF exported from a word processor is mostly text and already small — compression has little to work with. A scanned PDF is a stack of images and is where the big savings live. If your scan is enormous, combine downsampling with grayscale conversion for documents that do not need colour; the two together routinely cut a file by 80–90%.
Frequently asked questions
Does compressing a PDF reduce its quality?
It can, slightly, because image compression is lossy. Choosing a sensible level removes detail you cannot see at normal viewing size while keeping the document looking the same.
Why didn't my PDF get any smaller?
It is probably already efficient — typically a text-only document with no large images. There is little for a compressor to remove.
Is compression reversible?
No. Keep your original if you might need the full-resolution version later. Compression discards data permanently.
Will compressing blur my text?
No. Text is stored as instructions, not pixels, so it is unaffected. Only embedded images are recompressed.
How to tell whether your compression went too far
After compressing, spend a moment judging the result rather than trusting the size number alone. Zoom in to roughly the level at which the document will actually be read and look at the busiest image and the smallest text. If photographs show blocky squares or text edges look smeared, you have pushed too hard — step back to a higher-quality level. If everything still looks clean at normal viewing size, the data you removed was genuinely invisible and you have compressed well. Also compare the before and after sizes: a scanned document that barely shrank usually means it was already efficient or that its images were the kind a careful compressor declines to touch. There is no single correct setting, only the right setting for a given document and destination, so treat the first attempt as a measurement and adjust once. Keeping the original means you can always re-compress from full quality if your first choice was too aggressive.
Related guides
- How to Flatten a PDF (and Why You Might Need To)
- How to Convert a PDF to Grayscale
- PDF Compression Explained: DPI, JPEG Quality and Downsampling