Make your PDF smaller without breaking screen reader access.
Most PDF compressors strip embedded fonts, drop tags, or flatten the structure tree to save space. The result looks the same in a browser but a screen reader cannot read it. This tool compresses by recompressing images and re-packing the file. It never touches fonts, tags, alt text, the structure tree, bookmarks, language, or any other accessibility metadata. If your fonts would fail PDF/UA-1 anyway, the tool refuses to compress and tells you what to fix at the source.
No account needed. Most files compress in a few seconds.
Each finding pairs a verdict pill (BROKEN) with a recommended replacement. Re-export your PDF after replacing the font, then come back and compress.
The structure tree, fonts, alt text, and PDF/UA metadata in your file are unchanged. Compression touched only image XObjects and stream packing.
Most PDF compressors strip embedded fonts, drop tags, or flatten the structure tree to save space. The result looks the same in a browser but a screen reader cannot read it. Accessibility regulations (PDF/UA-1, WCAG 2.1 AA) and Indian rules (RPwD, GIGW 3.0) require the structure to remain intact.
This tool compresses by recompressing images and re-packing the PDF object stream. It never touches fonts, tags, alt text, the structure tree, bookmarks, language tags, or any other accessibility metadata. Before compressing, it runs the same font check as the Font Auditor. If any font would fail PDF/UA-1, the tool refuses to compress and tells you what to fix at the source.
Three guarantees, every time.
Before touching anything, the tool runs the same font check as the Font Auditor. If your PDF references Helvetica, Times-Roman, or any other Standard 14 PostScript font without embedding it, the tool refuses to compress and tells you exactly which fonts to replace at the source.
The tool decodes JPEG, JPEG 2000, and FlateDecode images, resizes anything larger than 2000 pixels in either dimension, and re-encodes as JPEG at quality 80. Images that have soft masks or stencil masks are skipped, because re-encoding to JPEG would lose the alpha channel they need.
Fonts, font programs, the structure tree, alt text, tag hierarchy, RoleMap, language attributes, MarkInfo, viewer preferences, bookmarks, annotations, and PDF/UA metadata are all preserved byte for byte. After compression, the file passes the same PDF/UA-1 checks as before.
Four steps, all on this page.
The file is held in memory. Nothing is written to our servers.
If any font is BROKEN, the tool refuses and tells you exactly what to fix in the program that made the PDF.
Images are recompressed and the file is re-packed. Every accessibility feature stays in place, byte for byte.
The compressed file downloads automatically. The result card shows you how much space was saved.
The tool refuses files that would fail PDF/UA-1 because of fonts that are not embedded. The fix is in the program you used to make the PDF (Microsoft Word, Adobe InDesign, LaTeX, LibreOffice, and so on). Replace the font (Helvetica becomes Arial or Source Sans 3, Times-Roman becomes Times New Roman or Source Serif 4) and re-export the PDF with font embedding turned on. Then come back and compress.
If you want a deeper diagnosis, run your file through the Font Auditor. It explains every font in the file, not just the ones that would block compression.
What we do with your file, and what the tool will not accept.
Four other tools, each one fits a different job. The Font Auditor is the natural prerequisite when this tool refuses your file.
A verdict for every font, with a recommended replacement when one is needed. Run this first if compression refuses your file.
AccessMitraDrop a PDF, get accessible PDF, Word, EPUB by email. For when you need the whole document fixed, not just compressed.
AltBridgeSend your subject expert a private link. They write a short note next to each figure. You get one CSV.
Document Accessibility CheckWord report, defects CSV, JSON summary, in under a minute.