Shrink one PDF or a batch in your browser. Pick a level, download in seconds. Files stay on your device. No signup, no upload, no watermark.
Drop PDFs here or click to browse
Up to 20 files · 100 MB each · PDF
Drag and drop or click to browse. Up to 20 PDF files, 100 MB per file.
Light keeps text selectable. Recommended balances size and quality. Maximum is for tight email limits.
Click Compress. One file downloads directly. Multiple files come back as a ZIP.
Compress PDF is a free online tool that reduces the file size of one or more PDFs directly in your browser. There is no signup, no upload, and no watermark. Your PDFs are processed locally on your device using two open-source libraries, pdf-lib and pdfjs-dist, and they never travel to a server.
Most PDF files are big because of the images inside them. A 12 megapixel photo from a phone weighs around 4 to 6 MB before it ever lands on a page. A 20-page proposal with one photo per page is heavy before you have written a single word. Scanned documents are worse: each page is a full-resolution image pretending to be a document. The text is not text, it is pixels.
A PDF is a container. Inside the container are images, fonts, vector drawings, and small bookkeeping records that tell a reader how to draw each page. Compression works on those parts in different ways. Re-encoding the images at a lower quality is the single biggest lever, often 60 to 90 percent of the total reduction. Dropping unused metadata, repacking object streams, and removing duplicate resources contribute the rest.
Light is the safest pick. Images are kept at original quality, text stays selectable, and embedded fonts are preserved. The win comes from repacking the file and stripping author history, annotations, form fields, and embedded resources. Expect up to 15 percent smaller. Use Light for legal documents, contracts, signed paperwork, and anything you may want to search later.
Recommended is the right pick for most PDFs. Each embedded image is downsampled to a screen-friendly resolution and re-encoded as a JPEG, while text, fonts, and vector drawings stay native. Your text remains selectable and searchable. Expect up to 50 percent smaller. Use it for slide decks with photos, reports with charts and images, scanned mixed-content documents, and screenshots-to-PDF.
Maximum goes harder on image resolution and JPEG quality. Images may show soft edges or slight JPEG noise when you zoom in, but text stays crisp and selectable. Expect up to 75 percent smaller. Use Maximum for proofs, drafts, and review copies where size matters more than zoom-in image detail.
The tool runs in any modern browser, so the same page works the same way on every operating system. On Windows or Mac, drag PDF files from File Explorer or Finder onto the dropzone. On a Chromebook, drag from the Files app. The browser handles the rest.
On iPhone and iPad, tap the dropzone and the iOS file picker opens. Pick PDFs from the Files app, iCloud Drive, Google Drive, Dropbox, or any other provider you have installed. On Android, the flow is identical through the system file picker.
Email providers cap attachments. Gmail rejects anything over 25 MB and silently routes you through Google Drive instead. Outlook caps at 20 MB on most plans. iMessage and WhatsApp accept larger files but compress them on your behalf, often producing a worse result than picking the level here. If you are sharing a PDF over chat or email, target under 20 MB.
Every byte of work happens on your device. The PDF is parsed in memory using open-source JavaScript libraries that ship with the page. Embedded images are decoded, downsampled, and re-encoded on a hidden canvas, then written back into the PDF in place. No part of your file is sent over the network. Closing the tab removes everything.
Yes. Completely free, forever. No signup, no watermark, no daily limit, no upgrade prompt. Use it as much as you want.
Yes. Text, fonts, and vector drawings stay native at every level. Only the embedded images get downsampled and re-encoded. Your text remains selectable and searchable in the compressed output.
The tool accepts files up to 100 MB. Files over 50 MB will show a slow-device warning. On phones and older laptops, very large files may use a lot of memory while compressing.
No. Every step runs in your browser using pdf-lib and pdfjs-dist. Your files never leave your device. You can confirm this by checking your browser Network tab.
Yes. The tool runs in mobile Safari, Chrome, Firefox, and Edge. Compression happens locally on the device, so no mobile data is used after the page loads.
Almost always because of images. Scans, photos exported as PDF, and slide decks with full-bleed photos all carry uncompressed or weakly compressed images. Text and vector graphics rarely take more than a few hundred kilobytes per file.
Not yet. Password-protected files are detected and skipped with a clear message. Remove the password in your PDF viewer first, then compress.
No. Every page in your input file is preserved in the output, in the same order.
Yes. Scanned PDFs usually shrink the most because the entire content is images. The tool detects scanned content and switches to a strategy tuned for it. A 50 MB scan can drop to 5 MB or less at the Maximum level. Use Recommended for archive copies you may want to read on a phone.
Yes. Drop up to 20 files. They are compressed in sequence using the same level. The result is delivered as individual downloads or one combined ZIP.
It depends on the content. PDFs with photos or screenshots see the largest cuts, often 30 to 70 percent at Recommended and up to 75 percent at Maximum. Text-only PDFs and already-optimized files may shrink very little since there is little image data to recompress. We never return a larger file than the original.
Open this page in any modern browser. Drag a PDF onto the dropzone, pick a level, and click Compress. The result downloads to your default download folder. No app to install and no Adobe Acrobat license needed.