Compress PDF by Re-encoding Pages
Reduce image-heavy PDFs by rendering and JPEG re-encoding pages. Choose DPI and quality, then compare the actual output size.
Drop files here. Files stay on your device.
Actual page re-encoding
This creates a new PDF from JPEG renders of the selected pages. Lower DPI reduces pixel dimensions; lower quality reduces JPEG size at the cost of visible artifacts. It can help large scanned or image-heavy files. Text-only or already optimized PDFs may grow, so compare originalBytes and resultBytes in the result.
The output keeps the visible crop and orientation as page images, with page dimensions measured in points. It does not retain searchable/vector text, layers, links, forms, bookmarks or original signatures. This is a deliberate raster compression method; retain your original if those features matter.
Quality and limits
Limits: 100 MiB input, 50 selected pages, 20 million pixels per page, 100 million pixels total and 100 MiB of encoded output. Pages render sequentially and processing can be cancelled. If the result is larger, increasing compression is optional; the tool never claims a guaranteed percentage saving.
Local processing and file limits
Files are processed in your browser; the PDF is not uploaded. The general limit is 100 MiB and 2,000 source pages, with the lower organizer/render/extraction limits described above. Encrypted or damaged files are rejected. Cancel, clear and input changes invalidate old results.
Questions & answers
Will text still be selectable after compression?
No. This mode saves page images. Use the original PDF for text selection or extraction; OCR is a separate task.
Are files uploaded to a server?
No. Files, input text and results stay in your browser. Required engines, fonts and support files load from this site.