PDFs have a tendency to balloon in size. A resume with a few high-res images easily hits 20MB. A scanned contract? 50MB+. When you need to email a file or upload it to a form, size limits become a real problem.

Most online PDF tools solve this by uploading your file to their server, compressing it, and sending it back. It's convenient, but it means your sensitive documents — contracts, resumes, financial records — leave your device and land on someone else's cloud.

In this post, I'll show you how to compress PDFs without ever uploading them to a server, including a browser-based approach I built for exactly this problem.

Method 1: Built-in OS Tools (Quick but Limited)

Windows

Right-click PDF → Print → Select "Microsoft Print to PDF" → Save.

This creates a compressed version, but you have zero control over the compression level. Sometimes it works; sometimes it over-compresses and blurs text.

Mac

Open in Preview → File → Export → Quartz Filter → "Reduce File Size".

Same limitation — no quality control.

Best for: Quick one-off compression when quality isn't critical.

Method 2: Browser-Based Processing (Privacy-First)

This is the approach I use for sensitive documents. Instead of uploading to a server, the PDF opens directly in your browser and gets processed locally with JavaScript.

How It Works Under the Hood

At the core is pdf-lib, a powerful PDF manipulation library that runs entirely in the browser:

import { PDFDocument } from 'pdf-lib';

async function compressPDF(file) {
  const arrayBuffer = await file.arrayBuffer();
  const pdfDoc = await PDFDocument.load(arrayBuffer);

  // pdf-lib doesn't have a direct "compress" flag,
  // but you can optimize by re-saving with reduced quality
  const pdfBytes = await pdfDoc.save({ useObjectStreams: true });

  return new Blob([pdfBytes], { type: 'application/pdf' });
}

For image-heavy PDFs, the real compression happens at the image level. Here's how I handle it in sotool.top:

// Extract images, compress them, re-embed
const pages = pdfDoc.getPages();
for (const page of pages) {
  const images = await page.embeddedResources();
  for (const image of images) {
    if (image instanceof PDFImage) {
      // Reduce image quality based on user-selected compression level
      const compressed = await compressImage(image, quality);
      // Re-embed compressed image
    }
  }
}

Key advantage: Your file never leaves your computer. The server never sees the content.

Trade-off: Browser memory limits. A 500MB scanned document might crash the tab, whereas a server-based tool could handle it.

For typical documents under 100MB, the speed difference is negligible.

A Real-World Test

I tested three approaches on a 15MB scanned contract:

The browser-based approach matched the server-based tool in quality, without the privacy trade-off.

When to Use What

Performance Considerations

Compressing PDFs in the browser has unique constraints:

Memory: Chrome caps each tab at ~1-2GB. For very large files, stream pages one at a time instead of loading the entire document:

// Instead of loading the full file
const pdfDoc = await PDFDocument.load(arrayBuffer); // Loads everything

// Process pages incrementally
const pdfDoc = await PDFDocument.load(arrayBuffer, {
  updateMetadata: false
});
// Only access pages you need

CPU: Image compression is CPU-intensive. For batch processing, use requestIdleCallback to avoid blocking the UI:

function processBatch(files) {
  files.forEach((file, index) => {
    requestIdleCallback(() => {
      compressPDF(file).then(updateProgressBar);
    });
  });
}

Web Workers: For heavy lifting, offload to a Web Worker so the main thread stays responsive:

// worker.js
import { PDFDocument } from 'pdf-lib';

self.onmessage = async (e) => {
  const { file } = e.data;
  const compressed = await compressPDF(file);
  self.postMessage({ compressed });
};

The Bottom Line

For sensitive documents, browser-based PDF compression is the sweet spot between convenience and privacy. Modern JavaScript libraries like pdf-lib make it entirely feasible, and the performance gap with server-based tools is shrinking fast.

If you want to try a browser-based tool without signing up:

👉 en.sotool.top/compress

Free, no signup, files never leave your browser.

What's your approach to handling large PDFs? Do you prioritize convenience or privacy? Let me know in the comments.