Compress PDF files without uploading them
Use this tool to compress PDF files that are too big to e-mail or too large for an upload form. You choose how: keep every page exactly as it is and let the file be stored more efficiently, or turn the pages into JPEG images for a much smaller file. Everything runs in your browser, so the PDF is not uploaded anywhere.
Two ways to reduce PDF size
| Keep text (lossless) | Smallest size (pages become images) | |
|---|---|---|
| Text | Stays selectable and searchable | Becomes part of a picture: no selecting, searching or copying |
| Links, bookmarks, form fields | Kept | Stop working |
| Typical saving | Small, sometimes large | Large for scans and photos |
| Best for | Contracts, reports, anything you will search or edit | Scanned documents, photo-heavy files, strict size limits |
Keep text (lossless)
The lossless method rewrites the file without changing what is on the pages. It removes page thumbnails (viewers draw their own), stores identical images or fonts only once when a file contains several copies, compresses data that was saved uncompressed, recompresses existing Flate data at the highest level when that is smaller, drops objects that nothing refers to any more, and packs small objects into compressed object streams. With Remove metadata ticked it also deletes the document properties, the XMP metadata packet and the private data some editing apps leave behind.
It does not touch JPEG photos inside the PDF, so the saving depends on how the file was made. A PDF exported by a modern office app is often already compact; files assembled from several sources, produced by older software or edited many times usually shrink more. For example, a three-page test report with the same image stored twice, one uncompressed page and a large XMP packet went from 345 KB to 167 KB, about 52% smaller, with its text still searchable.
Smallest size (pages become images)
This method draws every page, one at a time, at the resolution you pick and saves it as a JPEG. A new PDF is built with one image per page at the original page size, with rotation and cropping already applied, so it prints and displays like the original. The title is kept; other document properties are not.
The trade-off is real: text turns into pixels. A PDF that is mostly typed text can even get bigger this way, because text is stored very efficiently while an image of it is not. The tool never offers a bigger file: if the result is not smaller, it says so and keeps your original.
Compression levels
| Level | Resolution | JPEG quality | Use it for |
|---|---|---|---|
| Low compression | 150 DPI | 80% | Documents you may print |
| Medium compression | 110 DPI | 60% | Reading on screen, e-mail |
| High compression | 72 DPI | 40% | The smallest file that is still readable |
| Custom | 50–200 DPI | 10–95% | Your own balance |
Very large pages are scaled down so that no page needs more than 25 megapixels, which keeps the browser from running out of memory on posters and drawings.
Compress a PDF to 100 KB, 200 KB or any size
Application portals and government sites often accept only files under 100 KB, 200 KB or 1 MB. Choose Target size, pick one of these limits or type your own, such as 150 KB or 1.5 MB. The lossless method is tried first, and if it already fits, your text stays searchable. Otherwise the tool tries settings from 150 DPI at 80% quality down to 40 DPI at 25%, using a binary search so it needs at most four passes over the pages, and keeps the sharpest result under your limit. Sizes use 1 KB = 1,024 bytes, the way your computer reports file size.
If even the smallest setting is too big, you are told the smallest size this PDF can reach. Removing pages you do not need with Delete PDF Pages, or cutting the file into parts with Split PDF, are the usual next steps.
Why a PDF does not get smaller
- It is already compact. Lossless compression cannot shrink data that is already well compressed, such as JPEG photos.
- It is mostly text. Turning text pages into images adds size. Keep text instead.
- Images were huge to begin with. If you are building the PDF from photos, shrink them first with the image compressor; it is faster than compressing the finished PDF.
- It is password-protected. Encrypted PDFs are not changed, because a re-saved copy would lose its protection. Remove the password with Unlock PDF (you need the password), then compress the unlocked copy.
Privacy
The PDF is read and rewritten on your device with pdf.js and pdf-lib. Nothing is uploaded, and the compressed copy exists only in your browser until you download it. To strip personal details without changing anything else, use the PDF metadata remover.