How to compress a PDF, step by step

What the compression actually changes

Most of the weight of a large PDF sits in its images. Compress PDF re-encodes the embedded JPEG images at a lower resolution and quality, and rewrites the internal structure of the file more efficiently. Text and vector graphics such as logos, charts and lines are left alone, so they stay sharp and selectable.

The work happens in your browser. The file is not uploaded, there is no account, and no watermark is added.

Compress a PDF in five steps

The whole process takes a minute or two for a typical document.

• Keep an untouched copy of the original somewhere safe. The tool never overwrites it, but you may want to compare later.
• If the file is password-protected, unlock it first. Compress PDF cannot process protected files.
• Open Compress PDF, add the file and pick a preset. If you are unsure, start with Recommended.
• Run the compression and read the summary: it shows the original size, the new size and the percentage saved.
• Download the result, open it, and check a few pages at normal zoom before you send it.

Choosing a preset

Recommended is the balanced option and the right first attempt for nearly every document. It shrinks images noticeably while keeping photographs and scans comfortable to read on screen.

Maximum compression pushes image resolution and quality down further. Use it when you face a hard size limit, such as an upload form that rejects anything above a few megabytes, and when the images are not the point of the document. Expect softer photographs and, on scans, slightly fuzzier small print.

High quality keeps images closer to their original resolution. Choose it for portfolios, brochures, drawings or anything that will be printed, where you want some saving but cannot accept visible loss.

The convert pages to images option

In Maximum mode there is an extra choice: convert pages to images. It replaces each page with a picture of the page, which can shrink text-heavy scans further because the original structure and hidden layers are no longer carried along.

The price is significant. Text is no longer selectable or searchable, links stop working and the page cannot be edited. The tool shows a warning and never turns this on by itself. Use it only for documents that are already scans, and consider running OCR PDF afterwards to bring back a searchable text layer.

When the file barely shrinks

A PDF made mostly of text and vector graphics is already compact, so there is little for the tool to remove. The same is true of images that were already compressed hard. In those cases the saving may be a few percent.

If the result is not smaller than the input, the tool keeps your original and tells you the file was already well optimized. That is not an error; it means there is nothing safe to gain. To go further you have to change the content: fewer pages, lower-resolution source images, or a rebuilt document.

Other ways to get a smaller file

For a colour scan that only needs to be read or printed in black and white, PDF to Grayscale can reduce size. It renders each page to an image at 150 or 200 DPI and rebuilds the PDF, so text becomes non-selectable image content; treat it like the convert-to-images option and use OCR PDF if you need to search it.

If you are combining several documents, use Merge PDF first and compress the combined file once. Compressing the parts and then merging them works too, but it means more passes and more chances to leave one part at the wrong preset.

Check the result before you send it

Zoom in on the pages with the smallest print, signatures, stamps and photographs, since those show damage first. Confirm that the page count is unchanged and that any text you need to copy is still selectable.

If the quality is not good enough, go back to the original and run it again with a gentler preset rather than compressing the compressed file. Each round of image compression removes detail that cannot be recovered.