How to Compress a PDF to 100KB (And When You Simply Can't)

First, work out what kind of PDF you have

Everything depends on this and it takes five seconds to check. Open the file and try to select a line of text with your cursor. If a highlight appears over the words, you have a text-based PDF. If your cursor draws a rectangle across the page instead, you have a scan — a photograph of a page, stored as pixels.

The distinction matters because the two compress in completely different ways. A scanned page is mostly image data, and image data is where enormous savings live. A text-based PDF is already close to as small as its contents allow, because storing characters plus a font reference takes a tiny fraction of the space that storing a picture of those characters does.

A fifty-page contract exported from a word processor might be 200 KB in total. A single scanned page from a 600 DPI scanner can be four megabytes. Same information, twenty times the file size, and only one of them has any fat to remove.

If it is a scan: compress, and expect it to work

Scanned documents routinely compress by seventy to ninety percent with no meaningful loss of legibility, because almost all of their size is resolution nobody needs. A page scanned at 600 DPI contains four times the data of the same page at 300, and 300 is already beyond what any screen displays.

Work in this order, and stop as soon as you are under the limit:

• Start at a moderate compression setting, around 150 DPI. This is comfortably readable on any screen and prints acceptably.
• If you are still over, step up to the most aggressive setting, around 96 DPI. Small text will soften noticeably.
• If you are still over after that, the problem is page count rather than page quality — see below.

If it is a text PDF: compressing may make it bigger

This surprises people constantly and it is worth explaining, because it looks like a broken tool. Most browser-based compressors work by rendering each page as an image and rebuilding the document from those images. On a scan that is exactly right. On a text document it is actively counterproductive: you are converting compact text into a photograph of text, and photographs are bigger.

A well-behaved tool notices this and hands you back your original rather than the inflated version — which is why you may see a message saying the file is already optimised. That is not a failure. It means this particular method has nothing to offer your file.

It also destroys something. Once a page has been rendered to an image, the text is no longer selectable, no longer searchable, and no longer readable by a screen reader. For a document that was already a scan you lose nothing you had. For a text document you lose a great deal, which is the second reason not to compress files that were not large to begin with.

When compression alone will not get you there

If the most aggressive setting still leaves you above the limit, no amount of further compression will fix it gracefully. Squeezing a fifty-page scan into 100 KB produces something technically within the limit and practically unreadable, which fails the actual purpose of submitting it.

At that point you have three honest options, roughly in order of preference:

• Send fewer pages. Split the document and submit only what was asked for. This is almost always the right answer, and it is better for your privacy too — a form asking for three pages does not need the other thirty-seven.
• Rescan at a lower setting. Scanning at 200 DPI in greyscale produces a far smaller file than 600 DPI in colour, and for a text document the difference in legibility is negligible.
• Ask. Many forms state a limit that the system behind them does not actually enforce, and a great many organisations will accept the document by email if you explain the constraint.

Two habits worth forming

Keep the original. Compression is lossy and irreversible, and the compressed file is a copy made for one recipient with one limit. The next person who asks for the document will want the readable version, and you cannot recover it from a heavily compressed copy.

And do it in the right order when several steps are involved. If you are combining documents, merge first and compress once at the end — compressing each file separately and then merging pays the quality cost several times over for a worse result.

One last thing worth knowing: the files people compress under deadline pressure are rarely trivial. They are bank statements, identity documents, medical records, signed contracts. Doing this in a tool that runs in your browser rather than one that uploads means those documents never leave your device, which for this particular category of file is worth caring about.

Tools mentioned in this guide