How to reduce a PDF's file size (without ruining it)
A PDF that is too big to e-mail or upload is one of the most common document headaches. The good news is that most oversized PDFs can be shrunk a lot, and if you understand where the bytes come from, you can make them smaller without turning sharp pages into a blurry mess.
Why some PDFs are so large
A PDF is a container. It holds text, vector drawings, fonts, images and a structure that ties them together. Text and vector graphics are compact: a full page of typed text often needs only a few kilobytes. The big items are almost always something else.
- Images. Photos, screenshots and illustrations are usually the main reason a PDF is large. A single high-resolution photo can take several megabytes, and a brochure with dozens of them adds up quickly.
- Scans. A scanned document is not text at all. Every page is a picture of paper, so a 40-page scan is really 40 images. Scanners set to high resolution or colour mode produce especially heavy files, even when the pages are plain black-and-white letters.
- Embedded fonts. To look the same on every computer, a PDF can carry the fonts it uses. Usually only the characters actually used are embedded (a "subset"), which is small. But some files embed complete fonts, and fonts for Asian scripts can be large.
- Duplicated resources. When PDFs are merged or produced by certain programs, the same logo or font can end up stored many times instead of once. The pages look identical, but the file carries redundant copies.
- Leftover data. Some editors append changes to the end of a file rather than rewriting it, so old versions of content can linger. Metadata, thumbnails and hidden layers also add weight, though usually less.
Knowing this tells you something useful: if your PDF is a text-only report that is already a few hundred kilobytes, there is little to gain. If it is full of photos or scanned pages, there is often a lot.
Image resolution and file size
Image resolution in a PDF is measured in dots (pixels) per inch, or DPI: how many pixels of an image fall on each inch of the printed page. The relationship with file size is steeper than many people expect. Halving the resolution in both directions leaves only a quarter of the pixels. Going from 300 DPI to 150 DPI therefore removes roughly three quarters of the image data before any compression is applied, and going down to 72 DPI removes even more.
What resolution do you actually need?
| Resolution | Good for | Trade-off |
|---|---|---|
| Around 300 DPI | Printing, documents someone may zoom into, fine detail | Largest files |
| Around 150 DPI | Reading on screen, office printing, most e-mail attachments | Fine print in photos may soften slightly |
| Around 72 DPI | Quick previews, viewing on a phone or monitor at normal zoom | Visibly soft when printed or zoomed in |
Images are also compressed, for example with JPEG. Stronger JPEG compression shrinks files further but introduces artefacts: blocky areas, smudged edges and halos around text in scanned pages. Good compression tools balance both resolution and compression quality for you.
Vector text stays sharp
One important point: compressing a PDF that was created digitally (exported from a word processor, for example) mostly affects its images. The text is stored as characters and font outlines, not pixels, so it stays crisp at any zoom level even after strong compression. Scanned text is different. Because it is part of an image, it gets softer along with everything else when you lower the resolution.
Choosing a compression level
Most compression tools offer a few levels rather than dozens of settings. A practical way to choose:
- The file will be printed or archived. Use a light level that keeps images around 300 DPI. The savings are smaller, but nothing is lost that a printer would show.
- The file will be read on screen or printed on an office printer. A medium level around 150 DPI is usually the sweet spot. Most people will not notice a difference on a normal monitor.
- You just need it small. A strong level around 72 DPI produces the smallest file. It is fine for a quick look on a phone, but photos and scanned text will look soft when zoomed or printed.
When in doubt, start with the middle level, check the result, and only go stronger if you still need a smaller file. Always keep the original. Compression throws away image detail permanently, so you cannot get it back from the compressed copy.
Lossless optimisation
There is also a gentler kind of size reduction that does not touch image quality at all. Lossless optimisation rewrites the file's internal structure: it packs small objects together into compressed object streams, recompresses data streams more efficiently and removes some redundancy. Many optimisers also "linearize" the file, which lets a browser show the first page before the whole document has downloaded (often called fast web view).
The savings are usually modest, often a few percent, sometimes more for badly written files. It is a good choice when quality must stay exactly the same, such as for contracts or technical drawings, and it can be combined with compression.
Typical size limits
Limits differ by service and change over time, so check the specific one you are using. As a general guide, many e-mail services cap attachments somewhere around 20–25 MB, and the encoding used for attachments makes a file take up more space than its size on disk, so a file just under the limit can still be rejected. Online forms, job portals and government websites often have much lower limits, sometimes only a few megabytes per file. If a portal says "maximum 2 MB", aim comfortably below that.
Splitting as an alternative
Sometimes compression is not the right answer. If a document must stay at print quality, or it is so large that even strong compression will not get it under a limit, split it instead. You can send a long report as two or three parts, or send only the chapter someone actually needs. Splitting loses no quality at all. Our guide on how to split a PDF explains the options.
Another option is to remove what nobody needs: blank pages from a scan, duplicate pages, or an appendix full of photos. Deleting a few image-heavy pages can save more than any compression setting.
Checking the result
Before you send a compressed file, take one minute to check it:
- Compare the file sizes. If the new file is barely smaller, the PDF was probably text-based already, and compression will not help much.
- Open the compressed copy and zoom to 100% and 200% on a page with photos or small print.
- If the document is scanned, make sure small text, signatures and stamps are still legible.
- Check that the page count is unchanged and nothing is missing.
- If the result is too soft, go back to the original and use a lighter level. Do not compress an already compressed file again and again; each round degrades it further.
Be aware that rewriting a PDF can remove interactive features such as form fields, bookmarks or some links, and it always invalidates a digital signature. If a file has been signed, send the original or ask the sender for a smaller version.
Doing it on FlexyFiles
FlexyFiles compresses PDFs on its servers, so this works on a phone or an older computer too. Files up to 100 MB each are accepted.
- Open the Compress PDF tool and upload your file.
- Choose a level: "Light – print quality" (images kept around 300 DPI), "Recommended" (around 150 DPI) or "Strong – smallest file" (around 72 DPI, screen quality).
- Enter a name under "Save as" and click Run. The result appears in "Your files", where you can download it. Your original stays unchanged, so you can try another level.
If the compressed version would not be smaller than the original, FlexyFiles keeps the original size rather than giving you a bigger file. For a lossless pass, add the Optimize step in the full workspace at /pdf/app. There you can also chain steps, for example remove blank pages first with Delete pages, then compress, all saved as one result. See chaining steps in the workspace for how that works. If compression is not enough, use Split PDF to send the document in parts.