Why Is My PDF So Big? (And How to Shrink It)

You hit "save as PDF" or scan a document and end up with a 30 MB file when you expected 1 MB. PDFs have a reputation for being efficient, but a few specific patterns blow up the file size dramatically. Understanding which one is hitting your file determines how to fix it.

The four reasons a PDF is huge

1. The scanner saved at high DPI as PNG-per-page

Most scanners default to 300 DPI saved as lossless PNG. A single Letter-sized page at 300 DPI is 2550×3300 pixels. Saved as a PNG inside a PDF, that's roughly 2-3 MB per page. A 20-page scanned document = 40-60 MB.

The fix: re-render each page at a lower DPI (150 is plenty for screen reading; 200 for print) or as JPEG (lossy but tiny by comparison).

2. Embedded high-resolution photos

A "save as PDF" export from a Word document with embedded 12-megapixel photos preserves those photos at full resolution. Each photo takes 4-8 MB. A 10-photo report = 40-80 MB.

The fix: re-encode the embedded images at lower resolution and quality. The PDF still shows the same photos, just with less wasted detail.

3. Embedded fonts that aren't subset

PDFs can embed fonts to ensure the document renders the same on every device. If the embedding doesn't subset the font (only include the characters actually used), each font adds a few MB. A document with 5 different fonts can carry 15-20 MB of font data even if only ~200 characters are used from each.

The fix: re-export the PDF from the source program with font subsetting enabled. Most modern PDF generators do this by default; older ones don't.

4. Bloated metadata or duplicate streams

PDFs from some authoring tools carry duplicate copies of objects, large XMP metadata blocks, embedded thumbnails, or print-preview data. These rarely show up in your viewer but eat file size.

The fix: run the PDF through a "linearize and strip" pass that removes unused objects.

How to shrink a PDF

PDF Optimizer handles the first two cases by far the most common. The tool flattens each page to a JPEG at the DPI and quality you pick, then rebuilds the PDF. Result: typically 60-90% smaller for scan-heavy or photo-heavy PDFs, with minimal visible quality loss at the right settings.

Three presets:

• Light — typically 50% smaller. Best for already-clean PDFs where you want a small reduction with no visible quality loss.
• Medium — typically 70% smaller. The default. Visually identical to the original at normal viewing distances.
• Heavy — typically 85% smaller. Visible compression artifacts on close inspection but fine for sharing or web.

You can also pick a target file size and the tool will iterate quality settings to land under your target.

What flatten-based compression doesn't preserve

The trade-off: flattening a PDF to JPEG-per-page means text is no longer selectable in the output. The original PDF's text was vector-rendered glyphs; the flattened output is just images of pages.

For most "share this" use cases, this is fine. For documents you'll want to copy-paste from later, keep the original or re-OCR the flattened version with OCR PDF.

Quick decision tree

• Is the PDF a scan or photo-heavy report? → PDF Optimizer Medium preset.
• Is it a born-digital text PDF that's huge for no obvious reason? → It's probably font embedding. Re-export from source with subsetting on. If you can't, optimize anyway — text becomes images but you save space.
• Do you have unused pages bloating the file? → PDF Editor to remove them, then optimize.
• Are you sharing via email and need under 25 MB? → Optimize Medium first; if still too big, drop to Heavy.

Common gotchas

The optimizer didn't shrink my file at all. Some PDFs are already optimized — the photos are already JPEG, the fonts are already subset. There's nothing left to compress without going to extreme quality settings.

The output looks worse than the original. Try the Light preset, or use the custom-target mode to specify quality directly. The Medium and Heavy presets bias toward smaller files; Light biases toward higher fidelity.

The output PDF is bigger than the original. Rare but possible if the source was a vector PDF (small) and the optimizer flattened to JPEG per page (larger). The optimizer is best for image-heavy PDFs; for vector-heavy (text + line art) PDFs, the source format is already efficient.

Will it work on encrypted PDFs? Dropvert tries to load encrypted PDFs in permissive mode. If a password is required, unlock first with another tool.

Are my files uploaded? No. PDF Optimizer runs entirely in your browser via pdf-lib + pdf.js for rendering and re-encoding.