Make a PDF smaller in your browser: lossless cleanup, JPEG image recompression, or full page rasterizing, with the before and after size shown.
A PDF is big for one of two reasons: the images inside it, or the way its bytes are laid out. This tool tackles both in your browser, with three levels you pick yourself: Light rewrites the file without touching how it looks, Reduce image quality re-compresses the JPEG photos inside it while keeping the text selectable, and Rasterize turns every page into a JPEG for the biggest savings on scans. It shows the original and new size side by side, and it will not hand you a file that came out larger than the one you started with.
No. The file is read, rewritten and downloaded entirely inside your browser tab using pdf-lib and pdf.js plus your browser's own canvas JPEG encoder. Nothing is sent to a server, which matters for contracts, invoices and ID scans.
Light removes objects nothing points to, drops document metadata (title, author, creator) and re-saves the file with compressed object streams; it never touches the pictures. Reduce image quality does that plus decodes each embedded JPEG, scales it down to your chosen longest side and re-encodes it at your chosen quality; text and vector graphics stay as they were. Rasterize renders every page to a JPEG at your chosen DPI and rebuilds the PDF from those images.
Most PDFs from word processors are already well compressed: the text and fonts are tiny and compressed inside the file. In our test a 30-page text PDF went from 34.6 KB to 34.4 KB (0.6%) on Light, which is typical. Real savings come from images, so use Reduce image quality or Rasterize. A PDF that is mostly vector graphics or text will not shrink much at any level.
Rasterize replaces each page with a picture of that page, so the text, links and form fields are no longer real objects. That is the trade for the largest size drop, mainly on scans, which are pictures already. If you need searchable text, use Reduce image quality instead, or run the result through the OCR PDF tool.
Only plain JPEG images (what most photos in PDFs are). PNG-style images, images with transparency masks, CMYK and inverted-colour images are left exactly as they are, and the result panel tells you how many images were re-encoded and how many were skipped. If your PDF is full of non-JPEG images, Rasterize is the option that will shrink it.
You are told, and no download is offered, so you keep your original. This can happen with small or already-optimized PDFs, especially with Rasterize: a tiny 873-byte test PDF became 3.3 KB when rasterized.
Not directly: encrypted PDFs cannot be rewritten without the password. If you know the password, remove the protection with the Unlock PDF tool first, then compress the unlocked copy.
It depends on the content, so treat any single number with caution. On synthetic photo-heavy test files, Reduce image quality at 70% and 1600 px cut a 10.7 MB PDF to 1.67 MB, and Rasterize at 110 DPI cut it to 198 KB. Test photos made of random noise compress unusually well when scaled down, and real photos typically save less, so check the before and after sizes shown before downloading.
← All tools · Blog · Cheat sheets · FAQ