About the Compress PDF tool
This tool shrinks a PDF's file size using one of three approaches: a light, lossless-ish structural optimization that keeps text selectable, a strong compression that re-renders each page as a compressed image (the same technique most free image-heavy PDF compressors use), or a target-size mode that automatically searches for the quality level that gets you closest to an exact file size you set.
How it works
Upload a PDF and pick a mode. Light re-saves the PDF's internal structure more efficiently — safe, keeps all text and quality, but savings are modest and depend on the file. Strong renders every page to an image at your chosen quality level and rebuilds the PDF from those images. Target size lets you type an exact size (e.g. 500 KB or 2 MB) — the tool automatically tries different quality and resolution combinations and picks the closest result under or near that size. The tool shows the original size, new size, and percentage saved before downloading.
Assumptions and behaviour
- Light mode re-serializes the PDF with optimized object streams. It won't recompress embedded images or fonts, so savings vary and can sometimes be minimal.
- Strong mode rasterizes every page to a JPEG at the chosen quality/resolution preset, then rebuilds the PDF from those images page-by-page at the original page size.
- Target size mode binary-searches JPEG quality (and steps down resolution if needed) until the output is at or near your requested size — it can't always hit the exact number, especially for very small targets, but gets as close as possible.
- Strong and target-size modes are most effective on scanned documents or image-heavy PDFs — they can make a text-heavy, vector-based PDF larger in some cases, since rasterizing text is often less efficient than the original vector text.
- Strong and target-size modes remove selectable text and interactive elements (links, form fields, bookmarks) — the output is image pages only.
Limitations
- No OCR — text in a strong-mode output is not selectable or searchable, even if the source PDF's text was.
- Very large PDFs (many pages, high resolution) can take a while to process since everything runs on your device.
- Encrypted or password-protected PDFs can't be processed.
Privacy
The PDF is compressed entirely in your browser using pdf.js and pdf-lib. Nothing is uploaded or stored.

