You have a 40 MB PDF to email, but Gmail caps attachments at 25 MB. You compress it on some online site, and the result looks like a photocopy left in the oven. Frustrating.
Good news: compressing a PDF without visible quality loss is technically possible, as long as you understand what's happening under the hood and pick the right setting.
Where does PDF weight come from?
A PDF is a container that can host several content types:
- Vector text: ultra-light (a few KB even for 100 pages)
- Embedded fonts: from a few dozen to a few hundred KB per font
- Bitmap images (photos, scans): the #1 cause of heavy PDFs — often several hundred KB, sometimes several MB per image
- Metadata and structure: marginal
- Layers, forms, signatures: marginal
In the vast majority of heavy PDFs, images are what weighs. A 50-page text PDF usually stays very light; the same document with one high-resolution photo per page can easily exceed several dozen MB.
The 4 compression levers
1. Image resolution reduction (downsampling)
An A4 page printed at 300 DPI is sufficient for professional print quality. But on screen, 150 DPI is largely enough for reading.
Going from 300 to 150 DPI divides each image's pixel count by 4 (half the width, half the height) — and therefore, roughly, its weight — with no visible loss on screen.
2. Image re-encoding (JPEG vs PNG)
PNG scans (lossless) often weigh several times more than the same image as a good-quality JPEG. Good compressors re-encode photographic images as JPEG (quality around 80-85): the difference is usually invisible to the eye, and the gain is significant.
3. Font subsetting
A PDF with embedded fonts may integrate the entire font — including characters never used (Chinese ideograms, Cyrillic alphabets, math glyphs). Subsetting keeps only the actually-used characters.
The gain is counted in tens or hundreds of KB per font: marginal for one font, noticeable on a document that embeds several.
4. Removing unused data
PDFs often contain orphan objects: old page versions, software metadata, preview thumbnails, duplicated fonts. Rewriting the file with a tool like qpdf, or cleaning it with mutool clean -gg, removes those objects with no loss at all. The gain varies and is often modest.
Compression levels: when which?
| Use case | Recommended level | What to expect |
|---|---|---|
| Common email attachment | Medium | Clear gain, difference rarely visible on screen |
| Document to archive | Light | Modest gain, near-identical quality |
| Web or mobile display | High | Smallest file, fine on screen, mediocre in print |
| Professional print | None or very light | Print quality preserved |
The real gain depends mostly on the content: a photo- or scan-heavy PDF can shrink a lot, while a mostly-text PDF will gain almost nothing.
Many tools offer a single slider. A good tool exposes at least 3 levels: low (quality preserved), medium (compromise), high (web).
How to choose a PDF compression tool
Compression results vary widely from one document to another: a tool ranking based on a single file wouldn't tell you much about yours. The most reliable approach is to test for yourself, on your own document, with a few simple criteria.
What to check
- Where is the file processed? In your browser, or uploaded to a server? Your browser dev tools' "Network" tab shows you. If the file is uploaded, read the service's privacy policy and subprocessor list: hosting country, jurisdiction, retention period.
- How many compression levels? At least three (light, medium, strong), ideally with the result's size shown before you download.
- What happens to the text? Some tools keep vector text (selectable and searchable), others turn every page into an image. For a contract or report you need to search, that's decisive.
- What are the free-use limits? Maximum file size, files per day, possible watermark: these conditions change, so check them on the tool's website when you use it.
- Sign-up and ads? A compression tool doesn't need your email address to work.
How to judge the result's quality
- Open the original and the compressed file side by side, at 100% then 200% zoom, on a page with a photo and on a page with small text.
- Run a search (Ctrl+F or Cmd+F) in the compressed file: if it finds nothing anymore, the text has been turned into an image.
- If the document will be printed, print a test page.
- Compare the size you get with your actual need (for example your email provider's attachment limit): there's no point compressing harder than necessary.
Where PDFly fits
- Processing in your browser: the file is never sent to our servers.
- Three compression levels, with the before/after size displayed.
- Free, no sign-up, no ads; up to 100 MB per file, no daily cap.
- A limit to know about: PDFly re-renders each page as an optimized image, so the text is no longer selectable (see "Good to know" below).
How to compress on PDFly
- Go to pdfly.eu/en/tools/compress
- Drop your PDF (up to 100 MB — free, no sign-up)
- Choose the compression level:
- Light: maximum quality
- Standard (recommended): the best compromise, ideal for email
- Strong: the smallest file, ideal for web
- Click "Compress", then download the result
Everything happens in your browser. Your files are never sent to our servers.
Good to know: PDFly compresses by re-rendering each page as an optimized image. It is very effective on photo-heavy PDFs and scans, but the text of the compressed file is no longer selectable or searchable. For a text-heavy document where you need to keep selectable text, a local tool like Ghostscript (see below) is the better choice.
For very large files (over 100 MB)
PDFly's tools run in your browser, which means the whole document has to fit in your tab's memory. That's why each file is capped at 100 MB: beyond that, the page can freeze — especially on a phone. The limit protects your browser, not our servers.
For bigger PDFs (illustrated reports, massive high-res scans), two options:
- Split, then compress: cut the document into several parts with Split PDF, compress each part, then merge them back if needed.
- Compress locally with Ghostscript: free, open source, no size limit.
gs -sDEVICE=pdfwrite -dPDFSETTINGS=/ebook -dNOPAUSE -dQUIET -dBATCH \
-sOutputFile=compressed.pdf original.pdf
Common mistakes to avoid
- Compressing twice — each compression degrades a bit more. Restart from the original.
- Compressing an already-optimized PDF — gaining a few percent at the cost of slight quality loss is rarely useful.
- Compressing a scan at "High" level — text zones become blurry. Maximum "Medium" for scans.
- Compressing an electronically signed document — if the PDF carries a digital signature (with a certificate), any modification, compression included, invalidates it. Compress first, sign after.
- WeTransfer instead of compressing — for 1 file of 60 MB OK, but if you can compress it below your email provider's limit, you avoid third-party dependency (and keep the recipient in their email client).
In summary
Compressing a PDF without visible quality loss has been technically mastered for a long time. The secret isn't the tool — it's picking the right level.
For most emails and shares, the middle level is a good starting point: check the result on screen, and restart from the original with a lighter level if the quality doesn't suit you.
Compress a PDF now — free, in-browser, no signup.