Compress PDF

Reduce the size of a PDF in your browser and compare the result before downloading. Nothing is uploaded, so private documents stay on your device.

Drop a PDF here, or click to choose one

Your file never leaves this browser tab. Nothing is uploaded.

Honest note: pages are rebuilt as images, so text in the compressed file cannot be selected or searched. Best for scans and photo-heavy documents.

Make big PDFs small enough to send

Scanned documents and image-heavy reports routinely exceed email and portal limits of 5, 10 or 25 MB. This compressor rebuilds each page at a sensible resolution and shows you the before and after sizes, so you only download the result when it actually helps. Three levels let you trade quality for size: gentle for near-original clarity, balanced for everyday sending, and smallest for strict upload limits.

Before you compress

  • The compressed copy has image-based pages: text is not selectable or searchable.
  • Keep your original as the record copy; compress only the copy you send.
  • Documents with forms, signatures or fine print deserve the high quality setting and a quick visual check afterwards.

How compression works under the hood

This tool takes a different approach from simply repacking a PDF file. Your browser opens the document with pdf.js, the widely used open source renderer for viewing PDFs on the web, and draws each page onto a canvas at the resolution linked to the level you choose. High quality renders at one and a half times the normal page scale, balanced at a little above normal scale, and smallest below normal scale. Each rendered page is then saved as a JPEG image at a matching quality setting and placed into a new PDF built with pdf-lib, at the same physical page size as the original. Because the work is rendering and image encoding inside your browser tab, no server ever sees the document. That design explains both the strength and the trade off of the method. Scanned pages stored as enormous images shrink dramatically when re rendered at a sensible size, while the rebuilding step is also why text becomes part of the picture and bookmarks and form fields do not carry across.

Practical tips for getting better results

  • Start with balanced, then adjust. Balanced is the right first try for most scanned documents. If the result still misses a limit, run the original again at smallest rather than compressing an already compressed copy, which would blur pages a second time.
  • Match the level to the final use. A document read only on phones and laptops tolerates smallest well. A document that will be printed or read for small print deserves high quality, even if the saving is more modest.
  • Split before compressing a mixed document. If a 60 page file has only 5 photo heavy pages, use Split PDF to separate them, compress just that section, and rejoin everything with Merge PDF. The text pages keep selectable text and the file still gets much smaller.
  • Check a dense page at full size. After compressing, open the result and zoom in on the page with the smallest text, such as terms or footnotes. If those lines stay comfortable to read, the rest will be fine.

A worked example

Consider a scanned 12 page rental agreement saved at 18.6 MB, far above a letting portal limit of 5 MB. The pages contain black text, a few colour stamps, and two pages of property photos. At high quality, the rebuilt file comes out at about 7.9 MB, clearer but still too large. Running the original again at balanced produces about 3.8 MB, comfortably under the limit, with text that remains easy to read on screen. At smallest the file drops to about 1.9 MB, but photo pages look softer and small print needs a closer look. For this job balanced is the sensible choice to upload, while the original stays saved as the permanent record.

Common mistakes to avoid

  • Compressing a text only PDF and expecting a saving. A born digital report with embedded fonts may be only 220 KB to begin with. Rebuilding its pages as images can produce a file of 1 MB or more. The tool will warn you when the result is not smaller, and in that case the original is already the better file.
  • Deleting the original after compressing. Once pages are images, you cannot recover selectable text or original resolution from the compressed copy. Always keep the source file and treat the compressed version as a sending copy only.
  • Compressing twice to chase a limit. Each pass renders the previous pictures again at lower quality, so text edges get fuzzier every time. If one pass at smallest still misses the limit, splitting the document into two parts usually looks better than a second pass.
  • Ignoring interactive elements. Fillable forms, digital signatures, and clickable links do not survive rebuilding as images. Submit the original whenever the recipient needs to complete, sign, or click anything.

Frequently asked questions

How does this compressor shrink a PDF?

Each page is rendered in your browser at a reduced resolution and saved as an efficient JPEG image inside a new PDF. Most large PDFs are heavy because of oversized embedded images, so re-rendering pages at screen resolution removes most of the weight. You see the size before and after before deciding to keep the result.

Will my text stay selectable after compression?

No, and this matters. Because pages are rebuilt from rendered images, text in the compressed PDF is part of the picture and can no longer be selected, copied or searched. If you need searchable text, keep the original for records and use the compressed copy only for sending and sharing.

Why did my file barely shrink, or even grow?

Text-only PDFs are already tiny and efficient; turning their pages into images can make them larger. This tool shines on PDFs with photos and scans. If the result is not smaller, the tool tells you and you can simply keep your original.

Is my document uploaded for compression?

No. Rendering and rebuilding happen entirely in your browser. Your PDF never leaves your device, which is why this is safe for private documents.