How compression works here
Most of the weight in a big PDF is pictures: phone scans, photos and screenshots, often stored at far more pixels than the page needs. The tool reads the PDF, works out how large each image is actually printed on its page, and resamples it to the target resolution. The smaller image is saved as a JPEG and swapped in, but only if it really is smaller. Everything else in the file is left alone.
Pixels needed = printed size in inches × target DPI
Printed size in inches = size on the page in points ÷ 72
It also does three lossless clean-ups that never change how a page looks: it removes leftover objects that nothing in the document uses any more (common after a PDF has been edited several times), packs any data that was stored with no compression at all, and writes the file's internal index in a compact form.
Worked example: a phone-scanned lease
A phone scanner app saves each page as a 3024 × 4032 pixel photo, about 12 megapixels, placed on a Letter page of 8.5 × 11 inches. That is 4032 ÷ 11 = 367 pixels per inch, far more than a screen or office printer uses. At Balanced (120 DPI) the page needs only 8.5 × 120 = 1020 by 11 × 120 = 1320 pixels, so each page image drops to about one ninth of its pixel count. A 12-page lease that was 24 MB typically ends up in the 2 to 4 MB range, with the text on the scan still easy to read.
"Get it under a size": how the target is met
Choose a limit and the tool tries a ladder of ten settings, from 200 DPI at high JPEG quality down to 50 DPI at low quality. It checks the harshest setting first, so an impossible target is spotted straight away, then homes in on the gentlest setting that fits and confirms it with a real save. You get the sharpest file under the limit, not just the smallest one.
If even the harshest setting misses, the result says why: either the PDF has no images that can be re-saved, or its text, fonts and page structure alone are bigger than the limit. Only then is "flatten pages to pictures" offered.
What compression can't do
| Kind of PDF | What to expect |
|---|---|
| Scans and phone-scanned documents | Large savings, often 50 to 90%. |
| Reports and brochures with photos | Good savings on the photos; text pages stay the same size. |
| Text-only PDFs from Word, Google Docs or a bank | Little or no change. Text is already compact; there is nothing to throw away without turning it into pictures. |
| CMYK print files, JPEG 2000 and black-and-white fax scans | Those images are left as they are, and the result says how many were skipped. |
If a compressed file comes out no smaller, the tool keeps your original rather than handing you a bigger one. For files that will never fit, split the PDF into parts and upload them separately.
Privacy
The page downloads its code once, then reads, re-encodes and saves your PDF on your device, in a background worker so the page stays responsive. No file, page or password is uploaded, logged or kept. Load the page, switch off Wi-Fi, and it still works.
Questions people ask
Is my PDF uploaded to compress it?
No. The whole job runs in your browser: the PDF is read, its images are re-saved and the new file is written on your own device, then saved to your Downloads. Nothing is sent to a server, so it is safe for tax returns, medical records and signed contracts.
Will the text still be selectable and searchable?
Yes. This tool only re-saves the photos and scanned images inside the PDF. Text, fonts, vector drawings, links, bookmarks and fillable form fields are left exactly as they were. The one exception is the "flatten pages to pictures" option, which is only offered when nothing else gets under your size limit and which warns you before it runs.
Why did my PDF barely shrink?
Because there was little to shrink. Text and fonts are already stored compactly, so a PDF made from Word or Google Docs with no pictures usually saves only a few percent from the lossless clean-up. The big savings come from PDFs full of photos or scans, such as a phone-scanned lease or a photo-heavy report, which often get 50 to 90% smaller.
How do I get a PDF under 2 MB for an upload form?
Choose "Get it under a size", pick 2 MB and compress. The tool tries gentle settings first and steps down until the file fits, so you get the sharpest version that is under the limit. It aims under 2,000,000 bytes, which also passes forms that count 2 MB as 2,097,152 bytes. If it cannot fit with text kept, it tells you why and offers to flatten the pages.
Will the images look worse?
A little, in exchange for the size. Images are reduced to the resolution chosen (150, 120 or 96 dots per inch of printed size) and saved as JPEG. At 150 DPI photos and scans still print well; at 96 DPI they look fine on screen but soft on paper. Images already below the chosen resolution are not enlarged, and an image is only replaced when the new one is smaller.
Can it compress a password-protected PDF?
Yes, if you know the password. Type it when asked; it is used only on your device. The compressed copy is saved without the password or editing restrictions, so add protection again afterwards if you need it.