How to Compress PDF Size in 2026: Reduce 50-90% Without Losing Quality
The single most common question I get from students submitting assignments: "why is my 4-page PDF 60MB?" Nine times out of ten the answer is the same, they pasted phone photos straight into Word at full resolution before exporting. Compression is the cure, but the right setting depends entirely on what's bloating the file. Get that diagnosis wrong and you either fail to shrink it or you blur the one thing that needed to stay sharp. This guide walks the decision, not just the buttons.
Quick verdict: Open our free PDF Toolkit, click "Compress PDF," pick a quality level (High = lossless 10-30%, Medium = balanced 50-70%, Low = aggressive 70-90%), and download. Files process locally in your browser. Medium is the right default for email attachments; reach for lossless only when fine print or images must stay pixel-perfect.
What Makes a PDF File Big?
Knowing the cause helps pick the right compression approach.
| Content type | Typical size contribution | Compression potential |
|---|---|---|
| Embedded images (JPEG/PNG) | 50-90% of total | Very high — downsampling works well |
| Embedded fonts | 5-30% of total | Moderate — subset fonts |
| Scanned page images | 70-95% of total | Very high — lossy compression |
| Text and vector graphics | 5-20% of total | Low — already compressed |
| Form fields and metadata | 1-5% of total | Low — remove unused |
| Embedded attachments / video | Variable | Remove if unnecessary |
Lossless vs Lossy: The Core Tradeoff
Lossless compression (10-30% reduction)
Removes redundant data without changing any visible content:
- Deduplicates embedded fonts (one copy instead of many)
- Removes unused objects and orphan streams
- Recompresses already-compressed streams more efficiently
- Removes thumbnails and unused image data
- Subsets fonts to only used characters
Zero quality loss. Use this for legal documents, archival storage, anything where the output must look identical.
Lossy compression (50-90% reduction)
Reduces image quality to save space:
- Downsamples images (300 DPI to 150 DPI)
- Lowers JPEG quality from 95 to 75-85
- Converts PNG to JPEG when appropriate
- Removes color depth from images that don't need it
- Flattens transparency layers
Visible quality reduction in images. Use for email attachments, web sharing, mobile viewing — situations where screen-friendly quality is enough.
Choosing the Right Quality Level
| Level | Typical reduction | Image quality | Best for |
|---|---|---|---|
| High (lossless) | 10-30% | Identical | Legal, archival, print-ready |
| Medium | 50-70% | Slightly reduced, screen-fine | Email, internal sharing |
| Low | 70-90% | Noticeably reduced | Mobile viewing, archive thumbnails |
| Extreme | 85-95% | Heavily degraded | Only when size cap is mandatory |
Method 1: Browser-Based (Best for Most)
Browser tools compress PDFs locally using JavaScript. Files never upload.
Step-by-step using our PDF Toolkit:
- Open the PDF Toolkit in any modern browser.
- Click "Compress PDF" tab.
- Drag and drop your PDF or click to upload.
- Choose quality level (High / Medium / Low).
- Click "Compress." Processing happens in your browser.
- Compare before/after size shown in the result panel.
- Download the compressed file — the original stays untouched.
Most files complete compression in 5-30 seconds depending on size and content.
Method 2: Desktop Apps
Mac Preview (free, built-in)
- Open the PDF in Preview
- File → Export
- Format: PDF
- Quartz Filter dropdown → "Reduce File Size"
- Save
Free and fast but uses a single aggressive preset. Output is often 70-85% smaller with noticeable image quality loss. Good for one-off email compression.
Adobe Acrobat Pro
- Open the PDF in Acrobat
- File → Save As Other → Reduced Size PDF (or Optimized PDF for granular control)
- Choose compatibility (Acrobat 10+ usually fine)
- For Optimized PDF: adjust image downsampling, font subsetting, and discard options
- Save the compressed file
Most granular control. Best for production workflows where you need to balance specific quality factors.
PDF24 Creator (Windows, free)
Drag-drop "Compress PDF" tile, choose quality preset, download. Free and effective for casual use.
Method 3: Command-Line (For Automation)
Ghostscript
The most reliable command-line PDF compressor. Available on Mac (brew install ghostscript), Linux (apt install ghostscript), and Windows.
# Screen quality (most aggressive)
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 \
-dPDFSETTINGS=/screen -dNOPAUSE -dQUIET -dBATCH \
-sOutputFile=compressed.pdf input.pdf
# eBook quality (balanced)
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 \
-dPDFSETTINGS=/ebook -dNOPAUSE -dQUIET -dBATCH \
-sOutputFile=compressed.pdf input.pdf
# Printer quality (light compression)
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 \
-dPDFSETTINGS=/printer -dNOPAUSE -dQUIET -dBATCH \
-sOutputFile=compressed.pdf input.pdf
# Prepress quality (minimal compression)
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 \
-dPDFSETTINGS=/prepress -dNOPAUSE -dQUIET -dBATCH \
-sOutputFile=compressed.pdf input.pdf
The PDFSETTINGS presets correspond to:
/screen— 72 DPI, smallest size (aggressive)/ebook— 150 DPI, balanced (recommended default)/printer— 300 DPI, light compression/prepress— 300 DPI, color preserved, minimal compression
Advanced Compression Techniques
Convert scanned PDFs to searchable + compressed
Scanned PDFs are essentially image bundles. They compress well, but the real win is running OCR (Optical Character Recognition) and storing text alongside images. This makes the PDF searchable and dramatically smaller because the text layer replaces the image where readable.
- Tesseract (free command-line OCR)
- Adobe Acrobat's "Enhance Scanned PDF"
- ABBYY FineReader (best commercial OCR)
Replace embedded images with optimized versions
For technical workflows, extract embedded images, re-optimize them externally (jpegoptim, mozjpeg, oxipng), and reinsert into the PDF. Substantial savings beyond what built-in compressors achieve.
Subset embedded fonts
A PDF embedding a full font face (200KB) when only 30 characters are used wastes space. Font subsetting reduces this to just the characters needed.
Strip metadata and attachments
Author info, comments, embedded files, and revision history all add weight. Tools like qpdf can strip these:
qpdf --linearize --object-streams=generate input.pdf compressed.pdf
How Size Reduction Correlates With Quality
Realistic expectations for different content types:
| Content type | Lossless | Medium quality | Aggressive |
|---|---|---|---|
| Text-only document | 5-15% | 15-25% | 25-40% |
| Mixed text + images | 10-25% | 50-70% | 70-85% |
| Image-heavy document | 15-30% | 60-80% | 80-95% |
| Scanned document | 10-20% | 65-85% | 85-95% |
| Presentation export | 20-35% | 60-80% | 80-90% |
Pre-Compression Checklist
- Keep a copy of the original — compression is destructive in lossy modes.
- Identify the use case — email vs print vs web changes the right quality level.
- Check what's making the file big — if it's not images, lossy compression won't help much.
- Determine quality tolerance — can recipients tolerate visible image degradation?
- Try Medium first — if size is still too large, move to Low. If quality looks bad, try High instead.
- Verify text is still readable after compression — aggressive compression can blur fine print.
- Test on actual receiving device if quality matters — mobile screens hide compression artifacts that print reveals.
Strategy for Email Attachment Limits
Common email attachment limits:
- Gmail / Google Workspace: 25MB
- Outlook / Microsoft 365: 20MB (older) or 35MB (newer)
- Yahoo Mail: 25MB
- iCloud Mail: 20MB (Mail Drop for up to 5GB via link)
- Most corporate SMTP: 10-20MB
Compression strategy:
- Try Medium compression first
- If still too large, try Low
- If still too large, split the PDF and send in multiple emails
- For 50MB+ files, switch to file-sharing links (Google Drive, Dropbox, WeTransfer) rather than fighting attachment limits
When Not to Compress
- Legal documents going to court or regulators — quality and exact reproduction matter
- Print-ready files for commercial printing — printers need 300 DPI images, compressed PDFs at 150 DPI print blurry
- Documents with critical fine print — aggressive compression can blur text below 8pt
- Reference archives — keep originals for posterity; compress copies for distribution
- Documents with signatures — some lossy compression can affect signature integrity in some PDF readers
Common Issues and Fixes
- "File only shrunk 5%" — PDF is text-only with little compression headroom. Try font subsetting or remove unused resources.
- "Images look terrible after compression" — reduce compression level (try Medium instead of Low) or use a higher-quality preset.
- "Text became blurry" — aggressive image compression treated text as image. Use a less aggressive preset or use a text-aware compressor.
- "PDF won't open after compression" — tool produced invalid PDF. Try a different compressor; some legacy tools struggle with newer PDF features.
- "Compressed PDF is larger than original" — original was already optimized. Some "compress" tools re-encode and can occasionally increase size for already-compressed files.
The fix I teach before anyone touches a compression tool: solve the problem at the source. A student whose 4-page report comes out at 60MB doesn't have a compression problem, they have a "dropped six 12-megapixel phone photos into Word" problem. Compressing after the fact downsamples those images twice (once badly on insert, once on compress) and the result looks muddy. The better habit is to resize images to roughly 1500px wide before placing them, which alone takes that 60MB file down to 3-4MB with no visible quality loss, and then a Medium compress trims the rest. Compression is the last step in a clean workflow, not a rescue for a messy one. I've watched students go from "my file won't upload" to "done in 90 seconds" just by reordering those two habits.
Frequently Asked Questions
for pdf in *.pdf; do gs -sDEVICE=pdfwrite -dPDFSETTINGS=/ebook -sOutputFile=compressed-$pdf $pdf; done. Or use Adobe Acrobat's Action Wizard for GUI batch processing.The workflow I'd actually follow
Diagnose first: resize oversized images at the source, then run Medium compression for 50-70% reduction with screen-acceptable quality. Use browser-based tools for privacy and speed, desktop or command-line for batch jobs and granular control. Always keep the original, because compression is destructive and a digital signature won't survive it. The PDF Toolkit handles the common cases in under a minute per file.
