How to make a PDF smaller on a Mac

The built-in option, and its problem

Open the PDF in Preview, choose File then Export, and set the Quartz Filter dropdown to Reduce File Size. That is the whole procedure, and for a large scan it often works well.

The complaint is equally well known: on some documents it produces a file that is barely smaller, and on others it wrecks the images. Both happen for the same reason. The filter applies one fixed set of parameters regardless of what the document contains, with no quality slider and no preview.

Specifically, it downsamples images and re-encodes them at a low JPEG quality. If your PDF is mostly text there is almost nothing for it to act on, so the saving is trivial. If it contains photographs, they take a harder hit than most people would choose. The filter has no opinion about which case it is in.

Making your own filter

macOS lets you build a replacement, which is the fix most Mac users never hear about. ColorSync Utility — in Applications, Utilities — has a Filters tab listing the built-in Quartz filters.

Duplicate Reduce File Size and open the copy. Under Image Compression you get a quality slider, and under Image Sampling you get a resolution to downsample to. Setting quality to around 0.7 and sampling to 150 DPI produces something far more usable than the stock filter for a screen-read document; 200 DPI is a better floor if it may be printed.

Save it and the new filter appears in Preview's Export dropdown alongside the original. It is a fifteen-minute setup that permanently replaces the worst part of the built-in behaviour.

The command line route

If you have Ghostscript installed — via Homebrew, `brew install ghostscript` — you get finer control and can script it across a folder.

The relevant invocation sets a PDFSETTINGS preset: `/screen` for the smallest output, `/ebook` for a reasonable middle, `/printer` and `/prepress` for progressively higher quality. `/ebook` is the one worth trying first for a document that needs to stay readable.

This is genuinely useful for bulk work. It is also a dependency to install and a command to remember, which is a poor trade for compressing one file.

Check what you actually have first

Before any of this, a two-second test decides whether compression will help at all: try to select a line of text in the document.

• The text highlights — it is a text PDF. The size is coming from embedded images, if there are any. A text-only report will barely shrink, and that is the correct outcome rather than a tool failure.
• You get a selection box over an image — it is a scan. Every page is a photograph, compression will be dramatically effective, and you can expect to remove most of the size without a visible difference.

What to watch for whatever you use

The thing worth checking in the output is whether your text is still text. Some compressors — including some web tools — rasterise every page, which shrinks the file hard and turns your document into pictures of a document. Selectable text becomes unselectable, search stops working, and a screen reader gets nothing.

For a scan that costs nothing, because the pages were already images. For a contract or a report it is a real loss, and it is rarely disclosed. After compressing anything that matters, try selecting a line of text in the result. If you cannot, the tool rasterised it.

Frequently asked questions