How to Compress a PDF Without Losing Quality
A practical guide to shrinking PDFs for email and uploads — what actually gets compressed, why quality degrades, and how to pick the right setting for your file.
How to Compress a PDF Without Losing Quality
Email gateways cap attachments at 25 MB. Recruiting portals refuse anything over 5 MB. Government submission systems often have even stricter limits — sometimes as low as 2 MB. If you've ever been blocked from sending a PDF because of size, you know the frustration. This guide explains how PDF compression actually works, what trade-offs you're making, and how to get the smallest file with the least visual damage.
What's actually big inside a PDF
The first thing to understand is that PDF size is dominated by images. A 50 MB PDF is rarely 50 MB of text — it's usually 48 MB of pictures and 2 MB of everything else combined. Open a PDF in any inspection tool and you'll see the breakdown: photos, screenshots, scanned pages, and embedded charts are the heavyweights. The actual text content of a 200-page novel is well under a megabyte.
This matters because PDF compression is really image compression. When you "compress a PDF," what's happening under the hood is: every embedded image gets recompressed at a lower bit-depth, downsampled to a smaller resolution, or both. The text remains untouched. Embedded fonts are subsetted (unused glyphs removed). The cross-reference table is rebuilt for a slightly smaller index.
Knowing this changes how you think about results. A text-heavy PDF (a contract, a research paper, a printed book) won't compress much because there's not much image data to optimize. An image-heavy PDF (a slide deck full of screenshots, a scanned report, a photography portfolio) can compress 70–80% without obvious visual damage.
The three compression levels explained
Most online tools (including SwitchPDF's Compress PDF) offer three or four levels. Here's what each one actually does:
Low compression — images are downsampled to 300 DPI (the print standard) and recompressed at high JPEG quality (~85%). Visually, the result is indistinguishable from the original on any screen. File size typically drops 10–25%. Use this for documents you'll print or zoom into.
Medium compression — images are downsampled to 150 DPI at ~75% JPEG quality. Good for reading on screen at normal zoom. Slight softening becomes visible if you zoom to 200%+. File size drops 40–60%. The right default for sharing reports and proposals over email.
High compression — images are downsampled to 72 DPI at ~60% JPEG quality. This matches screen resolution exactly. Images look fine at fit-page zoom but obviously fuzzy if you zoom in. Compression artifacts (the blocky "JPEG halos" around text) become visible. File size drops 70–85%. Use only when you absolutely have to hit a strict size limit and screen-only viewing is the use case.
A practical workflow
Here's how I'd actually compress a real document.
Step 1: Try Medium first. Run the file through medium compression and check the result. For most documents this is the right balance. Open the output, scroll through, zoom in on a page with images. If it looks fine, you're done.
Step 2: If the file is still too big, try High. If Medium got the file from 50 MB to 12 MB but your limit is 5 MB, jump to High. Compare side-by-side with the original — does the soft text bother you? If not, ship it.
Step 3: If High still isn't small enough, split the document. Some compression jobs are unwinnable. A 200-page report with full-page photographs will be 30 MB even at High. At that point, use Split PDF to break it into smaller chunks and send them as separate files. Don't try to compress further — you'll just produce an unreadable result.
What compression won't do
Compression won't:
- Remove pages you don't want. Use Remove Pages or Extract Pages for that.
- Convert your PDF to a different format. Use PDF to JPG if you want a smaller image-only version.
- Help if the PDF was already optimized. Files exported from "Save as PDF" with the smallest-size preset are often already at the floor — running compression on them produces nearly identical output.
When compression goes wrong
The two most common compression failures:
The output is the same size as the input. This means the PDF was already optimized. There's nothing left to compress. Your only option is to remove pages or split the document.
Text becomes hard to read. This means you used High compression on a text-heavy PDF and the JPEG artifacts ate the letter edges. Re-compress at Medium or Low. Text never needs to be image-compressed unless your PDF is a scan — for scanned documents, run OCR PDF first to add a real text layer, then compress at Medium.
The bottom line
PDF compression is image compression. Pick the level that matches your viewing context: Low for print, Medium for email, High only when you must. Always check the output before sending. And if compression can't get you under the limit, splitting the document is almost always the right next move.
Related articles
When to Use OCR on a PDF (And When Not To)
OCR turns scans into searchable text — but it's overkill for many documents. Here's how to tell whether your PDF actually needs OCR and what to expect.
Online PDF Tools vs Desktop Software: Which Should You Use?
Adobe Acrobat costs $15+/month, free online tools are free. The real trade-off is more nuanced than price. Here's when each one is the right choice.
Markdown to Word: A Developer's Guide to Sharing Docs With Non-Developers
You write everything in Markdown. Your clients want Word docs. Here's how to bridge the gap cleanly with the right tools and a few formatting tricks.