Where a PDF's size actually comes from
In almost every large PDF, the bytes are in the pictures. Scanned pages, product photos and screenshots are stored as embedded images, often at far higher resolution than anyone will ever view or print. Text, by contrast, is stored as instructions and fonts, which take up very little room.
That is why this compressor works on the images. It finds the embedded JPEG images inside the file, re-encodes them at a lower quality and scales down the ones that are larger than they need to be, then repacks the document structure. Your text is never turned into a picture, so it stays crisp at any zoom level and remains selectable and searchable.
How to compress a PDF
Step 1: Add your PDF
Drop one PDF, up to 200 MB. It is read by your browser and never uploaded.
Step 2: Choose a compression level
Low for minimal change, Medium for the balanced default, or High for the smallest file.
Step 3: Compress
The tool rewrites the embedded images and repacks the document.
Step 4: Check the savings and download
Compare the original size, the new size and the percentage saved, then save the file.
What each level does
Low
Images re-encoded at 70 percent quality and capped at 1800 pixels on the longest edge. Closest to the original.
Medium
60 percent quality and a 1600 pixel cap. The recommended balance for sharing and email.
High
45 percent quality and a 1100 pixel cap, and the document title, author, subject and keywords are cleared.
Lossless repacking
Every level also repacks the file structure with object streams, which shrinks even text-only PDFs a little.
When to compress a PDF
Email attachments. Most mail servers reject messages over 20 to 25 MB, including the encoding overhead.
Upload forms. Portals for jobs, schools, tenders and visas often cap PDFs at a few megabytes.
Websites and downloads. Smaller brochures and manuals load faster for everyone.
Storage and archives. Scanned paperwork takes far less space once the page images are right-sized.
How much smaller will your file be?
It depends entirely on what is inside. Scans and photo-heavy documents often drop by 50 to 80 percent, because their images are the whole file. A text-only report, a generated invoice or an already optimized PDF may shrink only a few percent, since there is little to remove without rasterizing the text, which this tool deliberately never does.
If a file barely shrinks, that is usually good news: it means it was already efficient. The tool never returns a file larger than the one you gave it, so a document that cannot be improved is handed back unchanged.
- Scanned documents and photo brochures: large savings.
- Slide exports and reports with images: moderate savings.
- Pure text PDFs: small savings.
What is left untouched, and why
Only baseline JPEG images in RGB or grayscale are recompressed. Images in other encodings, such as CMYK for print, indexed color, fax-style CCITT or JBIG2 scans, and images that carry a transparency mask, are left exactly as they are. Rewriting those correctly is risky, and a broken image is worse than a larger file. Images under 8 KB are skipped because the saving would not be worth it.
Text, vector drawings, form fields, links and bookmarks are preserved. If you need a guaranteed small file from a scan that will not compress, convert the pages to images with the PDF to image converter and rebuild a PDF from those, accepting that the text will no longer be selectable.
Tips for smaller PDFs
- Try Medium first and only move to High if you still need to go smaller.
- Remove pages you do not need with the page extractor before compressing.
- For several files, use the bulk PDF compressor and download one ZIP.
- Unlock password-protected PDFs before compressing them.
Compress PDF FAQ
Is my PDF uploaded to a server?
No. The file is read, rewritten and saved inside your browser, which is why it works with confidential documents.
Will the text get blurry?
No. Text stays vector and selectable. Only embedded images are re-encoded, so blur can only appear in photos at high zoom.
Why did my PDF hardly shrink?
It is probably mostly text, or its images are already optimized or in a format that is deliberately left untouched.
Does compression remove metadata?
The High level clears the document title, author, subject and keywords. Low and Medium leave them in place.
What is the file size limit?
One PDF up to 200 MB per run.
Can it compress a password-protected PDF?
Remove the password first. Encrypted files cannot be rewritten reliably.