← All guides

Convert a PDF to PDF/A

The court portal, the land registry, the university archive: "PDF/A only." It's the profile of PDF built to still open identically in 2050: every font embedded, color pinned to an embedded ICC profile, no scripts, no encryption. pdfa converts with one command, and checks its own work.

How to do it

pip install pdfblah
pdfblah doctor --install     # fetches Ghostscript, which performs the conversion

pdfblah pdfa contract.pdf -o contract-a.pdf

converted to PDF/A-2b  ->  contract-a.pdf
  output intent: yes  |  XMP PDF/A id: yes
  converted with an embedded sRGB output intent; for formal certification, validate with veraPDF

What makes this conversion the proper one

• A real output intent. Naive Ghostscript runs skip the embedded sRGB profile, producing files that claim PDF/A but fail validators on the first check. pdfa wires the profile in.
• Verified, not assumed. The output is re-opened and inspected for the PDF/A markers (output intent, XMP identification) before success is claimed.
• Honest about scope. Formal certification is a discipline of its own; the reference validator is veraPDF, and the report says so instead of overpromising. What you get reliably is what "PDF/A only" portals are asking for.
• --level 1, 2 (default), or 3 for the PDF/A part your recipient names.

Tips

• Encrypted PDFs can't be archival by definition; unlock first (the error says exactly that).
• Archiving a whole case folder? pdfa is a batch recipe step: sanitize, then pdfa, over everything.
• Scanned documents convert fine; run OCR first so the archived file is also searchable, which archives appreciate.