Z

Compress PDF

Reduce PDF file size in your browser by rebuilding the document. No upload required.

Runs in your browser — files never leave your device

How it works

A PDF is internally a collection of hundreds or thousands of numbered objects — pages, fonts, images, metadata — plus an index that says where each one lives. Files accumulate structural fat over time: many generators write that index as a verbose plain-text table, leave small objects uncompressed, and — most wastefully — save edits by appending to the file, so every revision of an edited object stays in the file forever alongside its replacement.

This tool parses the document with pdf-lib and writes a clean, single-generation copy using object streams: the small bookkeeping objects are packed together into flate-compressed streams, the cross-reference index is written as a compact compressed stream, and only the current version of each object is kept. A typical report that has been through several rounds of annotate-and-save carries every earlier version of its changed objects; the rebuilt file contains just the live ones.

Just as important is what it does not do: images and fonts are copied through byte for byte, never re-encoded or downsampled. That makes the operation lossless — pages render identically, text stays selectable — but it also bounds the savings. A file dominated by scanned page images will barely shrink, because the image bytes are untouched; the win comes on structure-heavy documents, and the before/after readout tells you honestly what happened, including when there was nothing to remove.

Practical notes:

  • Useful when a file just misses an attachment limit and you don’t want any quality loss.
  • Digital signatures are invalidated by any rewrite — compress before signing, not after.
  • Password-protected PDFs can’t be loaded; save an unprotected copy first.

The whole rebuild happens in your browser’s memory. The PDF is never uploaded, so nothing confidential ever leaves your machine.

Frequently asked questions

What does the compression actually do?
It parses the PDF and writes a clean new file with a more efficient internal structure: the many small bookkeeping objects are packed into compressed object streams, the cross-reference index becomes a compact compressed stream instead of a plain-text table, and superseded object versions left behind by earlier edits are not carried over. It’s structural housekeeping, not lossy compression.
Does it recompress or downsample the images inside?
No. Image and font data are copied through byte for byte — nothing is re-encoded, resized or reduced in quality. That’s what makes the process lossless, and also what limits how much an image-heavy file can shrink.
How much smaller will my PDF get?
It depends entirely on how the file was built. PDFs with long edit histories or old-style uncompressed structure can shed a noticeable share of their size, while files from modern, well-behaved exporters often only lose a few percent — and a document that is mostly scanned images will barely move. The before/after readout shows you the real number for your file.
Why did my file barely shrink?
Because its bulk is image or font data, which this tool deliberately leaves untouched. In a scanned document, the page photographs are often well over 90% of the bytes, so restructuring the remainder saves little. Shrinking such files further requires downsampling or re-encoding the images, which is a lossy operation this tool doesn’t perform.
Will the PDF look or behave differently afterwards?
Pages render identically, because the content that describes them is copied through unchanged — text stays selectable and links keep working. One exception to know about: a digitally signed PDF loses its signature validity, since any rewrite of the file changes the bytes the signature was computed over.
Can it compress password-protected PDFs?
No — encrypted files fail to load and you’ll see an error. Open the PDF with its password in a viewer, save an unprotected copy, and compress that instead.
Is my PDF uploaded to a server?
No. The file is parsed, rebuilt and re-saved by the pdf-lib library running entirely in your browser — it never leaves your device, and no copy exists anywhere else at any point. That makes it safe to use on contracts, statements and other documents you wouldn’t hand to a random website.

Related tools