Compress PDF

Start with lossless. If the file is already well-built — most PDFs exported from Word, LaTeX, or a design tool are — expect only a few percent, and that is the honest ceiling for a lossless rewrite. Switch to target-size mode when a form imposes a hard limit you must clear, or when the document is a scan where the text is already an image and there is nothing to lose.

Enter the number the portal asks for and the tool searches downward through render scale and JPEG quality until the output fits, then reports the size it achieved. If even the lowest settings overshoot — usually a document with too many pages — it tells you instead of handing back a file that will be rejected. Removing pages you do not need with Delete Pages or Split PDF first is usually the fastest way to close the gap.

Because each page becomes a JPEG, the output has no selectable text, no working links, no form fields, and no OCR layer. Screen readers cannot read it. That is an acceptable trade for a scanned certificate you need to upload, and a bad trade for a contract someone needs to search. Keep the original either way — the tool never modifies the file you dropped in.

Scanned documents are a stack of photographs, which is why they balloon to several megabytes per page. They are the ideal case for target-size mode: the text is already raster, so re-encoding costs nothing conceptually and the savings are large. If you need per-page control instead, export with PDF to Image and compress each page in Image Compressor.

Both modes run entirely in your browser — parsing with pdf-lib, rendering with a self-hosted copy of PDF.js. Your PDF is never uploaded, nothing is logged about its contents, and closing the tab discards it. Document metadata such as author and producer is stripped in both modes.

No registration, no daily quota, no watermark on the output. Compress as many PDFs as you need, at any size your browser can hold in memory.