Why is my PDF file so large?

A ten-page text document should be somewhere between 50 and 300 KB. If yours is 40 MB, something specific is inside it — and it is almost always one of four things. Working out which one you have is the difference between a fix that takes ten seconds and a fix that ruins the document.

1. It is a scan, so every page is a photograph

This is the single most common cause. A scanned or phone-photographed page is not text. It is a full-colour image of text, and at 300 DPI an A4 page is roughly 2480×3508 pixels. Saved as colour JPEG that is typically 1–3 MB per page; saved as PNG or uncompressed TIFF inside the PDF it can be 10 MB or more per page.

Twelve scanned pages at 3 MB each is a 36 MB PDF containing about 4 KB of actual words. The giveaway: you cannot select the text with your cursor, and the file size scales almost exactly with the page count.

The fix is to re-encode those page images at a lower resolution — 150 DPI is still comfortably readable on screen and roughly quarters the size. That is what compress PDF to 2 MB and its 5 MB and 1 MB siblings do when a structural re-save is not enough.

2. iPhone and Android scans are worse than desktop scanners

Phone scanning apps — including the Notes app and Files — usually capture at full camera resolution and apply little compression, because they are optimising for legibility rather than size. A five-page phone scan routinely comes out larger than the same document scanned on an office machine.

If the app offers a quality or "file size" setting, drop it before scanning. If the PDF already exists, re-compressing it afterwards is the only option left.

3. Canva, Illustrator and InDesign exports carry the originals

Design tools embed placed images at their source resolution, not at the size they appear on the page. A 20-megapixel stock photo cropped into a 3 cm square in your layout is still a 20-megapixel photo inside the exported PDF. Add a few of those and a two-page flyer becomes 60 MB.

Transparency and blend modes make it worse: when the exporter cannot preserve them structurally it flattens the affected area into a raster image, sometimes at very high resolution. This is why a Canva PDF with a drop shadow behind a logo can be several times the size of the same design without it.

Where you have the source file, exporting again with image downsampling turned on is by far the better fix. Where you only have the PDF, compress it.

4. Embedded fonts, and why they are usually innocent

Fonts get blamed constantly and rarely deserve it. A subset embed — only the characters actually used — is typically 20–80 KB per font. A full embed of a large font family can reach a few megabytes, and a CJK font can be 10 MB or more, which is the one case where fonts genuinely dominate a file.

If your PDF is 40 MB and contains Latin text in three fonts, the fonts are not your problem. Look at the images.

How to tell which one you have in about five seconds

• Try to select a line of text. If you cannot, it is a scan — cause 1 or 2.
• Divide the file size by the page count. Over roughly 1 MB per page with no photographs on those pages points at scanned images.
• Is the size concentrated in a few pages? Split the PDF and check. If two pages account for 90% of the file, those pages have big placed images — cause 3.
• Open it and count the photographs. If there are genuinely twenty high-resolution photos, the file is large for an honest reason and compression will visibly cost you.

Split PDF is useful for that third test, and it is also a legitimate fix in its own right — if the recipient only needs pages 3 to 7, sending 5 pages instead of 60 beats compressing all 60.

What compression actually costs

There are two different operations hiding behind the word "compress". The first is a structural re-save: removing duplicated objects, unused resources and revision history. It is lossless and sometimes gets you 10–30% for free. The second is re-encoding the embedded images at lower resolution, which is where the real savings are and where quality goes.

Our PDF tools try the structural pass first and only rasterise if that misses the target — and they tell you which one happened, because "your 40 MB file is now 1.9 MB" means something very different depending on the answer. Small text in a downsampled scan is the first thing to suffer; check a page of body text at 100% zoom before you send it.

If it just needs to fit an email

Gmail's attachment ceiling is 25 MB and Outlook's is 20 MB, but the practical limit is lower because attachments are base64-encoded in transit, which adds about 33%. A file sized to exactly the limit can still bounce. Aim under it: 10 MB is a comfortable target for almost any mail system, and 4 MB for portals that are stricter.

All of these run entirely in your browser. The PDF is not uploaded, which for scanned contracts, bank statements and medical records is the point.