Compress PDF to a Target Size

A PDF is not one single blob of data. It is a collection of numbered objects, each holding a different kind of content, tied together by a cross-reference table that tells a reader where to find each one. Understanding which of those objects is eating your megabytes is the whole game:

Most oversized PDFs are oversized for one of a few very specific reasons, and each has a different remedy:

• Images at far higher resolution than needed. A page scanned or exported at 600 DPI holds four times the pixel data of the same page at 300 DPI, and roughly sixteen times that of 150 DPI. On a screen, or printed on an office printer, most of that detail is invisible. This is where the large savings come from.
• The same image stored over and over. A letterhead logo placed on every page of a 200-page report can be saved as 200 separate copies of identical data. Storing it once and referencing it 200 times costs nothing in quality.
• Data written without efficient packing. Plenty of programs that produce PDFs write their objects out loosely, leaving ordinary compression on the table. Repacking that data changes nothing you can see.
• Whole fonts embedded for a handful of characters. Reducing an embedded font to only the glyphs the document actually uses can save hundreds of kilobytes in a document with several typefaces.
• Accumulated leftovers. Files that have been annotated, signed and re-saved repeatedly carry history and discarded objects that no longer render.

The first, third, fourth and fifth of those cost you nothing at all - the document looks exactly as it did. Only reducing image detail is a genuine trade, which is why it is worth trying everything else first.

This single distinction predicts your result better than any setting:

A quick way to tell them apart: open the PDF and try to select a sentence with your mouse. If the text highlights word by word, it is digital-born. If you can only draw a box over it like a picture, it is a scan.

Some compressors reach a small file the crude way, by flattening every page into a picture. That number looks good and the document is quietly ruined: the text can no longer be selected, searched, or read aloud by assistive technology, and anything that later needs to extract data from it has nothing to work with.

• Pages are never turned into flat images. Text that was selectable before is still selectable afterwards.
• The page count never changes. A 240-page document comes back with 240 pages.
• Nothing is re-encoded needlessly. A file already at or under your target is returned untouched, so it costs you no quality at all.

We also try the optimizations that cost no quality before touching any image detail, so a bloated file can often reach your target while still looking identical to the original.

Pick the limit you actually have to satisfy rather than the smallest number possible - every megabyte you give back is detail you keep. Common real-world limits:

• 100 - 300 KB - government, university and visa application forms, which are often the strictest
• 500 KB - 2 MB - job applications, recruitment portals and most web upload forms
• 5 MB - a safe size for email attachments that has to survive the recipient's mail server too
• 10 - 25 MB - client portals, document management systems and tender submissions

If a target cannot be reached without wrecking the document, you get the smallest usable version we can produce along with a note saying so, rather than a file that technically hits the number but is unreadable.

The Portable Document Format was created by Adobe in 1993 and published as an open standard, ISO 32000, in 2008. Its purpose is fidelity: a PDF describes a page precisely enough that it renders identically on any device, which is why it became the default for contracts, invoices, forms and anything that has to be archived or printed exactly as intended.

That same fidelity is the reason file sizes vary so widely. The format makes no attempt to decide how much detail your images need - it faithfully stores whatever it was given, whether that is a lean export from a word processor or a 600 DPI colour scan of a page that is only black text on white paper.