Why Scanned PDFs Are Large
A scan stores a photograph of each page. High pixel counts, color noise and lightly compressed images can dominate the file size. Already optimized scans may become larger after another encoding pass.
PDF Compression vs Image Compression
This workflow analyzes each document before rendering supported scan pages and encoding them as JPEG images in a new PDF. It does not optimize arbitrary PDF objects or preserve editable structures. No page is silently omitted.
Before Compressing a Scan for an Upload Form
Check whether the recipient needs selectable text, a valid signature or an archival original. This compressor creates a new image-only copy, so keep the original and use it only for eligible scans. If the upload limit is strict, compare the complete downloaded PDF's byte count with that limit; a smaller preview image is not proof that the document will fit.
Need fewer pages instead? Extract the required PDF pageswithout treating the whole document as a scan. To join documents, useMerge PDF. For selectable text, usePDF to Text; none of these tools adds OCR to a scan.
Choosing Resolution and Quality
PPI determines output pixels per inch; JPEG quality determines lossy image encoding. Lower PPI can erase tiny print and fine table lines. Compare before and after at zoom, using the same settings as the final output. Preview page JPEG bytes exclude PDF overhead; the complete size is measured only after compression.
Scanned PDF vs Text-based PDF
A page with image painting and no detected text or interactive structures may be a scan. No extracted text alone is not proof. Vector paths, masks, unfamiliar painting instructions and small isolated images produce Unknown / Review Required and remain blocked in this release. Mixed documents are not partially converted.
Does Compression Remove OCR Text?
Raster rebuilding can remove an invisible OCR layer along with links, forms, annotations, bookmarks, signatures and special metadata. This tool blocks detected text operators, interactive structures and signed PDFs. It performs no OCR and does not restore those features. Explicit consent is required even for likely scans.
Privacy
The original PDF stays unchanged on your device. Browser workers process one page at a time, clean temporary canvases and image bitmaps, and release their parsers. Cancellation stops local work. Filenames, scan images and extracted content are neither uploaded nor recorded in analytics.
