A 60 MB PDF is a small disaster: it bounces off email limits, takes forever on hotel Wi-Fi, and makes the person on the other end quietly resent you. The good news is that most large PDFs are large for fixable reasons — and you can usually cut them by 70–90% with no visible quality loss.
Why file size matters
Email gateways routinely reject attachments over 10–25 MB. Document management systems choke on bulk uploads. And every extra megabyte is latency for whoever opens the file on a phone. Smaller PDFs aren't about neatness — they're about the document actually reaching its destination and opening quickly.
There's also a cost angle: if you process thousands of documents through an API or store them at scale, a 5× size reduction is a 5× storage and bandwidth saving.
What actually makes a PDF large
Before you compress anything, it helps to know where the weight is. In practice, almost all of it comes from a short list:
- Full-resolution images. A phone photo dropped into a page can be 300 DPI when the page only needs 150.
- Embedded fonts. Entire font families embedded when only a handful of glyphs are used.
- Scanned pages. Each page stored as a giant lossless image instead of selectable text.
- Redundant objects. Duplicated images, leftover revisions, and unused form data.
If your PDF is mostly text but still huge, it's almost always scanned images. Run OCR first, then compress — you'll often go from 40 MB to under 2 MB.
Lossless compression first
Always start with lossless optimization: stripping duplicate objects, subsetting fonts, and re-compressing streams. This shrinks the file with zero change to how it looks — there's no reason not to do it.
For many office documents (contracts, reports, slide exports) lossless alone gets you a 30–50% reduction. If that's enough to clear the email limit, stop here — there's no quality trade-off to second-guess.
Downsampling images safely
When lossless isn't enough, the next lever is image resolution. The trick is matching DPI to how the document is actually used:
| Use case | Target DPI | Typical size cut |
|---|---|---|
| Screen / email only | 96–120 DPI | 80–90% |
| Office printing | 150 DPI | 60–75% |
| High-quality print | 300 DPI | 20–40% |
Downsampling is lossy, so do it last and eyeball the result. For anything heading to a professional printer, keep 300 DPI and lean on lossless plus font subsetting instead.
Compress with imisspdf
Putting it together, here's the workflow that works for almost every file:
- Open Compress PDF and drop your file in.
- If it's a scan, run OCR PDF first to convert images to text.
- Pick a level: Recommended for email, Strong for web, Light for print.
- Download, and confirm the text still looks sharp before you send.
The whole thing takes a few seconds, files are deleted within an hour, and nothing is used to train a model. That's the entire process — no Acrobat license required.