How to Compress a PDF Without Wrecking the Text
Why PDFs get huge, which part of the file is actually costing you megabytes, and how to shrink one for email or an upload form without turning the text to mush.
Updated August 11, 20265 min read
A PDF that will not fit through an email gateway is almost never big because of its text. Text is tiny — an entire novel is a few hundred kilobytes. PDFs get large because of what is embedded alongside the text: scanned page images at camera resolution, embedded fonts, and revision history from whatever produced the file. Knowing which of those is inflating your document tells you exactly which lever to pull, and stops you from destroying the text in pursuit of savings that were never there.
First, find out what kind of PDF you have
There are two fundamentally different documents that both end in .pdf, and they compress in opposite ways.
| Type | How to tell | Where the size is |
|---|---|---|
| Digital PDF | You can select and copy the text | Fonts, vector graphics, embedded images |
| Scanned PDF | Text is not selectable — it is a picture | Page images, almost entirely |
Try to select a line of text in your reader. If the cursor highlights words, it is a digital PDF. If it draws a rectangle over a picture, every page is an image and your file is essentially a photo album with a PDF wrapper. That single test determines your whole strategy.
Compressing a scanned PDF
This is the easy case and the one with the biggest wins. A phone-scanned page is typically captured at 3000+ pixels across — around 300 DPI for A4, which is print resolution. Nobody reading your document on screen needs that. Dropping to 150 DPI cuts the pixel count by 75% and remains perfectly legible on any monitor and on most printers.
The second win is colour. A page of black text photographed in colour stores three channels of data to represent something that is essentially two-tone. Converting scanned text pages to greyscale routinely takes another 40–60% off, and for pure text documents the visual difference is nil. On our compressor, greyscale and a custom DPI are Pro fine-tuning; on Free, Balanced (the default) takes scanned pages to 150 DPI and Strong to 96 DPI.
Compressing a digital PDF
Here the text costs you nothing and there is far less headroom. The savings come from three places, in descending order of impact.
- 1Embedded images. A report with twenty full-page photographs pasted in at camera resolution is carrying twenty photos it displays at a fraction of their size. Downsampling those to the size they are actually shown at is usually the entire saving.
- 2Font subsetting. A PDF can embed complete font families — every weight, every glyph, including alphabets your document never uses. Subsetting keeps only the characters present in the file and can save several megabytes on font-heavy documents.
- 3Object cleanup. Editing software leaves behind orphaned objects, old revisions, unused form fields and thumbnails. Removing them is pure win with no visual cost.
If your digital PDF is large and contains no images, fonts are almost certainly the reason. If it is large and full of charts exported from a spreadsheet, those charts may have been embedded as high-resolution bitmaps rather than vectors — re-exporting them as vectors from the source is the real fix.
Runs entirely on your device, so contracts, medical records and financial statements are never sent to our servers.
Compress a PDFWhen compression is the wrong answer
Sometimes the document is simply too big for the channel, and squeezing it further only makes it unreadable. Three alternatives usually beat further compression:
- Split it. If the recipient only needs the appendix, send the appendix. The PDF splitter extracts a page range into its own file, and the page remover drops pages you never needed to send.
- Send a link instead. Most email gateways cap attachments around 25MB. Cloud storage has no such limit and gives you the ability to revoke access later — something you cannot do with an emailed attachment.
- Convert pages to images. For a one-page document that only needs to be *seen*, a PDF-to-JPG conversion produces a file a fraction of the size that opens instantly on a phone.
What compression cannot fix
Compression reduces size. It does not remove content, and this trips people up in a way that has real consequences. If your PDF contains a page you blacked out with a rectangle, compressing the file does not remove the text underneath — the text is still there, still selectable, still recoverable. Covering something is not deleting it. Genuinely removing content requires proper redaction — the PDF editor's Redact tool, a Pro feature — which deletes the underlying objects rather than drawing over them.
Metadata is a separate question. Our compressor's Balanced, Strong and Flatten modes remove the title, author and editing history by default (Lossless keeps them); the name of the software that made the file stays. Many other compressors leave all of it in, so if the document is going to someone outside your organisation, run it through the metadata editor and check what it is telling them about you.
A practical checklist before you send
Check the page count
Are you sending pages the recipient does not need? Removing four scanned pages beats compressing all twenty.
Identify scanned vs digital
Try to select text. This tells you whether to expect a 90% saving or a 20% one, and stops you blaming the tool for the difference.
Compress, then open it
Actually read the result at 100% zoom before sending. Compression settings that look fine on a title page can render body text unpleasantly on page nine.
Check metadata and hidden content
Especially for anything leaving your organisation. Tracked changes, comments and author fields all travel with the file.
For the full set of document tools — merging, splitting, page numbering, watermarking and form filling — the PDF tools hub collects them in one place with notes on which to use when.
Frequently asked questions
How much can a PDF realistically be compressed?
Scanned documents commonly shrink by 80–95%, because they are photographs stored at far higher resolution than anyone views them. Digital text-based PDFs usually only give up 10–30%, since the text itself is already tiny — most of the saving there comes from embedded images and fonts.
Will compressing a PDF make the text blurry?
Not in a digital PDF — text is stored as vectors and is unaffected by compression. In a scanned PDF the text is part of an image, so aggressive downsampling can soften it. Staying at 150 DPI or above keeps scanned text sharp on screen and in most printing.
Is it safe to compress confidential PDFs online?
Only if the tool does not upload them. Our PDF compressor processes the file on your own device, so contracts and medical records are never transmitted to a server or stored anywhere.
Why is my PDF still large after compressing?
Most often because it is a digital PDF with little image content, where there was never much to save. Check whether the file contains embedded fonts or high-resolution charts. If it is genuinely mostly text and still huge, it may carry revision history — re-saving through a clean export from the source application often helps.