What Compress PDF does
PDFGrover's Compress PDF tool reduces a PDF's file size by re-encoding the images embedded inside it. Text, fonts and vector shapes are preserved — pages are never rasterised or flattened. Embedded photos and scans are down-sampled to a target resolution and re-compressed, which is where almost all the size saving comes from in real documents. That's why a text-only PDF barely shrinks while an image-heavy brochure can drop by 70–80%.
When to compress a PDF
- Email attachment limits — get a file under a 10 MB, 5 MB, or strict 1 MB ceiling without deleting content.
- Upload forms — bank portals, visa/government sites, and job boards often reject anything over 1–2 MB.
- Faster sharing and storage — smaller files send, sync, and back up quicker, and use less cloud quota.
- Slimming scans — scanned documents are usually huge and compress dramatically with little visible loss.
How to compress a PDF
- Upload a PDF (up to 500 MB) — drag it in or click to browse.
- Pick a level — Low, Medium, High, or Very High. Medium is the recommended default; choose Very High only when you must hit a hard size limit.
- Click Compress and download the smaller file when it's ready.
The four compression levels
Each level sets an image DPI and a JPEG quality, nothing more. Pick based on where the output file is going to end up.
- Low — 200 DPI, quality 85. Best for documents that will be printed or archived. Images stay sharp on paper and on large screens.
- Medium (recommended) — 150 DPI, quality 65. A good default for email attachments and most web sharing. Images remain crisp on screens at normal zoom.
- High — 110 DPI, quality 45. Good for quickly emailing a document that just needs to be readable. Photos look slightly softer at full zoom but text and embedded screenshots stay clearly legible.
- Very High — 50 DPI, quality 20. Use when you must hit a strict upload-size ceiling (bank portals, government forms with a "max 1 MB" rule). Photos will look soft; embedded text-images become legible-but-blurry; native PDF text remains sharp because text is vector, not image, data.
You can also toggle grayscale. Converting colour images to B&W trims another ~5 % off the final size on most documents and is a good fit for scanned text documents that don't need colour.
What size reduction to expect
In our testing on image-heavy PDFs, the average output sizes are approximately:
- Low → about 25 % smaller than the original
- Medium → about 50 % smaller
- High → about 65 % smaller
- Very High → about 80 % smaller
These are averages. Your file will land inside a range around those numbers depending on what's actually in it. A PDF that's already optimised — or one that's mostly text with no embedded photos — will compress far less, because there's little image data left to squeeze. On a 5 MB / 317-page text-heavy share-issue prospectus we benchmark internally, Very High lands at about 28 % reduction; on an image-heavy brochure of the same size, Very High typically clears 70 %. Small files under 500 KB generally only shrink a little; the overhead of fonts and structure is a bigger fraction of their total size.
When compression won't help much
- Text-only PDFs. Fonts are already compressed in the source file; there's no image data to re-encode.
- PDFs already run through a compressor. A second pass gains little and can soften images further if you pick a lower level.
- Scanned PDFs where each page is one big JPEG at 300 DPI. Those compress a lot on Medium / High, but you may prefer our PDF to grayscale tool first if the scan is a colour photocopy of black text.
Encrypted PDFs
If your PDF is password-protected, the tool asks for the open password before compressing. The password is used only to decrypt the working copy for Ghostscript to read — it is never stored and never written to the output file.
After compressing
- Still too big? Re-run at a higher level, or enable grayscale. Past Very High there's little left to gain — the remaining size is fonts and structure, not images.
- Need to email several files? Merge PDF them into one document first, then compress once — usually smaller than compressing each separately.
- Mostly a colour scan of black text? PDF to Grayscale before compressing often beats compression alone.
- Sharing page images instead? PDF to JPG can be smaller still when you only need pictures, not a document.
Troubleshooting
- Barely shrank — the PDF is text-heavy or already optimised; there's little image data to re-encode. That's expected, not a failure.
- Text looks fine but photos are soft — that's the trade you chose; drop to Medium or Low for sharper images at a larger size.
- "Password required" — enter the open password so the file can be read; it's never stored or written to the output.
- Upload rejected — confirm it's a valid PDF under 500 MB.
Privacy and file handling
Compression runs on our secure server, not in your browser, because this kind of image re-encoding can't run client-side. Your uploaded PDF is written to a temporary folder, processed, and deleted as soon as your download is ready. Close the browser tab mid-run and the job is cancelled and the temporary files cleared automatically. No sign-up, no watermark, and no copy kept for any other purpose.