A file that will not send is a familiar problem. Upload your PDF above and it comes back smaller — still a normal PDF, still selectable text, just without the weight most documents carry unnecessarily.
How much smaller depends entirely on what is inside, and the honest answer is worth knowing before you start. Below: where the megabytes actually live, why some files barely move, and what to do when compression is not enough.
One rule explains almost every PDF size question: images are the weight, text is not.
The complete text of a 200-page report might occupy 300 KB. One photograph straight from a phone occupies 4 MB. A scanned page — which is a photograph of text rather than text — can run to 1 MB on its own. So a 40-page scan is a 40-megabyte file, while a 400-page text document is a small one.
Compression works by re-encoding those images at a resolution suited to screens rather than print, and by cleaning up structural waste: duplicated resources, unused objects, and font data for characters the document never uses.
Which means results split into three predictable groups:
Being straightforward about this matters, because "same quality" is a claim worth qualifying.
Text is untouched. It is stored as instructions to draw glyphs, not as pixels, so it stays perfectly sharp at any zoom and remains fully selectable and searchable. Compression does not go near it.
Images are re-encoded, and that is lossy. Detail is discarded — chosen to be detail the eye is least likely to miss, but discarded nonetheless. At normal reading size on a screen the difference is usually invisible. It becomes visible if someone zooms deeply into a photograph, or prints the document at full size on good paper.
It cannot be undone. Decompressing does not restore what was removed. Keep your original as the master copy and send the compressed version — not the other way round.
Three cases where you should check the result before sending: a document going to a commercial printer, technical drawings with fine lines and small labels, and photographs a client will examine closely. Everything else is safe.
Most mail systems cap attachments somewhere around 20–25 MB. The trap is that attachments are encoded for transmission, and that encoding adds roughly a third to the size in transit.
So a 22 MB file arrives at the server as something closer to 30 MB and bounces, despite looking comfortably under the limit. Aim for 15 MB or less if the file must go by email, and remember the recipient's server has its own limit which may be stricter than yours.
Upload forms are often tighter still. Government portals, job application systems and court filing services commonly cap at 5 or 10 MB, and some reject anything over 2.
Three situations where something else solves the problem better.
You only need part of the document. Extracting the relevant pages with Split PDF removes their weight entirely rather than reducing it. Sending one page instead of twenty is both smaller and better for privacy, since the rest never leaves your computer.
The scan includes blank pages. Feeders routinely capture the backs of single-sided sheets, and each blank scan still carries a full-size image. Delete Pages removes them completely.
You expected cropping to help. It does not. Cropping sets a viewing rectangle and leaves the content outside it inside the file, so the size barely changes. Crop for appearance, compress for size.
Scans are where compression helps most, and there is a useful ordering to know.
If you also intend to make the document searchable, run OCR first, then compress. OCR reads the page images to recognise characters, so it performs better on the higher-quality originals. Compressing first gives the recognition engine less to work with.
Worth knowing too: OCR adds an invisible text layer without removing anything, so it makes the file slightly larger. Compressing afterwards more than covers the difference.
For future scans, check your scanner settings. Many default to 600 DPI in full colour, which is right for photographs and wildly excessive for a printed page. Around 300 DPI, in greyscale for text documents, produces a file a fraction of the size with no practical loss.
Still too large? Split PDF extracts only the pages you need, and Delete Pages removes blanks from a scan. OCR PDF makes a scan searchable before you compress it, and Merge PDF combines documents once they are all lean. Browse all PDF tools.