How to Reduce PDF File Size

Why PDF files get large

A PDF is a container. Text is stored as character codes plus the fonts that draw them, and drawings as lines and curves, so documents made of text and simple graphics are usually compact. Large files usually come from one of these:

• Oversized images. An image keeps all its pixels even when it is shown small. A full A4 page at 150 dpi needs about 1240 × 1754 pixels, and many phone cameras take photos with several times that many.
• Scans. Each scanned page is a picture of the paper, not text. A colour A4 scan at 300 dpi is roughly 2480 × 3508 pixels per page.
• Embedded fonts. PDFs carry their fonts so pages look the same everywhere. Many programs embed only the characters used, but a fully embedded font with a large character set, such as Chinese or Japanese, can outweigh all the text in the document.
• Duplicated resources. Merged PDFs each bring their own copy of the fonts and images they use, so the same font or logo can end up in the file several times.
• Metadata, thumbnails and editing data. Document properties, an XMP metadata block, page thumbnails, and data that some design programs save so they can reopen the file for editing. Most of this is small, but saved editing data can be large.
• Incremental saves. Some programs save changes by appending them to the end of the file, so older versions of everything that changed stay inside. A PDF edited and saved many times can carry a lot of dead weight.

Two ways to shrink a PDF

Lossless clean-up rewrites the file without changing how any page looks. In Compress PDF this is the Keep text (light) method. It packs the PDF’s internal objects into compressed object streams, compresses data that was stored without compression, deletes objects nothing refers to, and removes page thumbnails, editing-app private data and metadata (the document title is kept). Because the whole file is written out again, old versions left behind by incremental saves are dropped too. Text stays selectable and searchable, and links keep working.

The light method does not re-compress images or merge duplicate fonts and images, so scanned and photo-heavy files usually shrink very little. It helps most with files that were saved without compression or edited many times.

Re-rendering pages as images is the Convert pages to images (strong) method. Each page is drawn as a JPEG picture at the resolution and quality you choose and placed on a new page of the same size. Only that picture is kept, so duplicate fonts, oversized photos and hidden data disappear, and scans often shrink a lot. The cost is real: text can no longer be selected, searched or copied, links and form fields stop working, and screen readers cannot read the result.

Use the light method for anything people need to search, copy from or fill in, and the strong method for scans and documents that only need to be looked at. Keep your original either way.

Choosing resolution and JPEG quality

The strong method offers three resolutions, 72 dpi (smallest, screen only), 100 dpi (balanced) and 150 dpi (sharper, larger), plus a JPEG quality slider from 30% to 90% in steps of 5. It starts at 100 dpi and 60%.

Resolution sets how many pixels each page gets. PDF pages are measured in points, 72 to the inch, so at 72 dpi one point becomes one pixel: an A4 page becomes an image of about 595 × 842 pixels, and 10-point text is only about 10 pixels high. At 150 dpi the same page has more than four times as many pixels. Quality sets how hard JPEG squeezes those pixels. Low settings leave smudges and blocky patches around sharp edges, which show first on text.

These are rules of thumb, so always check the result at the size people will view it:

• For reading on screen or uploading to a portal, start with the defaults and zoom in on the smallest text before sending.
• If the file is still too large, try 72 dpi before pushing the quality very low. It suits large, clear text, not footnotes or fine print.
• Use 150 dpi for small print, detailed drawings, stamps or signatures that must stay legible.
• For printing, a common rule of thumb is 300 dpi, which this tool does not offer. Keep documents that will be printed as text with the light method, or reduce the images where the file is made.
• Change one setting at a time and click Compress PDF again. Each run starts from your original file, so repeated attempts do not degrade the result.

Send only the pages you need

The easiest bytes to save are the ones you never send. Upload forms often want one statement period, one certificate or the signed page of an agreement, not the whole document.

The PDF Splitter copies the pages you choose into a new PDF without re-rendering them, so text and image quality are unchanged. Pick Page Range (for example 4-6) or Select Pages, tick Save the selected pages together as one PDF and click Extract Pages. The new file contains only the resources its pages use, so it is usually much smaller, though not always in proportion: a page that uses an embedded font still carries that whole font.

If the extracted pages are still too big, compress them next. The strong method then only has to render the pages you are keeping.

Fix the size where the PDF is made

If you control how the PDF is created, fixing the size there usually gives a smaller and better-looking file than compressing it afterwards.

• Scanning. For ordinary paperwork, 200 to 300 dpi is a common rule of thumb; higher settings mostly add size. Choose greyscale when colour carries no meaning, since colour stores three values per pixel and greyscale one. Black and white is smaller still for typed text, though stamps, pencil marks and photos can look rough.
• Word processors. In Microsoft Word for Windows, select a picture, choose Compress Pictures, untick Apply only to this picture and pick a lower resolution before saving as PDF. The Minimum size (publishing online) option when saving as PDF also gives a smaller file than Standard. Many other programs have similar image settings in their PDF export options.
• Photos. Shrink photos before they go into a PDF. The Image Compressor re-saves one image at a time as JPEG or WebP, with a quality slider and an optional maximum width. Choose JPEG, then combine the images with JPG to PDF, which accepts JPG and PNG and embeds JPEG files unchanged, so they are not compressed twice. Avoid PNG for photos: it is lossless and usually much larger (see JPG vs PNG vs WebP).
• One export instead of many merges. Each exported PDF embeds its own fonts, so exporting a combined document once is usually smaller than merging several separate exports.

When the result is not smaller

Compress PDF shows the Original and Result sizes side by side with the percentage change. If the result is not smaller, it suggests keeping the original but still offers Download the result anyway. The usual reasons:

• Light method, image-heavy file. Images make up most of the size, and the light method leaves them alone. Try the strong method.
• Strong method, text document. A page stored as text is tiny. The same page as a JPEG is a picture full of sharp edges, which JPEG compresses poorly, so the result can be larger than the original. Keep the original.
• Settings too high. At 150 dpi and high quality, the new page images can outweigh a scan that was already well compressed. Lower the resolution first, then the quality.
• Already efficient. If neither method helps, send fewer pages or rebuild the PDF from smaller source files.