Compress PDF Free Without Losing the Quality

Choose "Reach a target size", set the target to 500 KB — either with the preset button or by typing an exact figure — and compress. You do not have to guess at a compression level: the server tries progressively stronger settings and keeps the gentlest one that fits under your number, so no more quality is given up than the target actually requires. The result tells you the final size and which setting it settled on. If even maximum compression cannot get there, you are told plainly rather than handed a file that misses the limit, and you still get the smallest version we could produce. The next answer covers what to do in that case.

Because of what is inside it, and target mode will tell you when you have hit that wall rather than quietly missing your number. Text, vector graphics, and form fields are already compact, so a 40-page contract may only be 300 KB to begin with and has almost nothing left to remove. Scanned pages are the opposite: each page is a full photograph, and a 300 DPI colour scan is roughly 1-2 MB per page before compression. If a scanned document will not reach your limit, you are up against the resolution the scan was captured at — the pixels have to go somewhere. Splitting the document with Extract PDF Pages and sending the parts separately is usually faster than fighting it.

The slider maps directly onto two things: the resolution images are downsampled to, and the JPEG quality they are re-encoded at. At 0 it holds images at 300 DPI and quality 92. At 50 it targets 150 DPI at quality 70 — still fine for screen reading and ordinary printing. At 100 it goes down to 72 DPI at quality 40, which is screen-only territory where photographs visibly soften. Real text is not touched at any setting, because it is stored as font instructions rather than pixels, so a text-heavy document stays perfectly sharp even at maximum. The rule of thumb: 150 DPI is the point below which printed output starts to look worse than the original.

It can, and how much is up to you. Every PDF also gets lossless treatment that costs nothing visually — streams are re-deflated harder and the internal objects are repacked — and on a text-only file that is often the whole saving. The lossy part only applies to raster images, and only where it actually helps: an image is re-encoded solely if the smaller version really is smaller, and is left untouched otherwise. Because you choose the intensity and can run the file again at a different setting, you can find the point where it is small enough and still looks right instead of accepting whatever a fixed one-setting compressor hands back.

Because compressing it saved less than 3%, so you are told rather than handed a "compressed" file that is effectively identical. This is normal for PDFs exported from modern software, files that have already been through a compressor, and text-only documents with no images to downsample. There is nothing wrong with the file — there is simply nothing meaningful left to remove, and the honest answer is more useful than a fake saving.

Gmail and most consumer mail providers cap attachments at 25 MB, Outlook.com at 20 MB, and many corporate mail servers at 10 MB. Government, university, and job-application portals are far stricter and commonly ask for under 2 MB, 1 MB, or 500 KB per upload. Check the limit you are actually being held to before compressing — clearing 25 MB is a very different job from clearing 500 KB, and aiming at the wrong one either wastes quality or wastes a second attempt.

Up to 100 MB per file. That covers the large scanned documents and image-heavy reports that are the usual reason someone needs a compressor in the first place.

No. Compression discards image data permanently, so the compressed file cannot be restored to its original quality. Keep your original: compress a copy, send the copy, and hold on to the full-quality file. If a compressed export turns out to be too soft, go back to the original and export again at a lower setting rather than trying to recover the compressed one.

No. Real text stays selectable and searchable, internal and external links keep working, and form fields stay fillable. Compression targets the images and the redundant structural data in the file, and leaves the interactive and textual layers alone.

Usually yes, and black-and-white scans are protected automatically. Pure mono or bilevel scans — the output of most document scanners set to text mode — are skipped by the image compressor entirely, because JPEG re-encoding wrecks that kind of image rather than shrinking it. Colour and greyscale scans do get downsampled, so those are the ones to watch: stay in the middle of the slider and check the smallest text in the preview, not the headings, since fine print degrades first. Palette-based and CMYK images are also left alone, for the same reason.

No. The exported file comes back clean — no watermark, no logo, and nothing added to any page by the tool.

It is completely free with no signup, no credit card, and no daily cap. Open the page, upload a PDF, compress it, and download the result.

Your file is uploaded over an encrypted connection, used only to produce the compressed version, and removed after your session ends. It is not stored long-term, sold, or shared. Compression happens on our server rather than inside your browser, so an internet connection is required.