open sourceApache-2.0v0.1.1

HTML in.
PDF out.

Every PDF job your product needs, in one small binary — a CLI and an HTTP API that do exactly the same things.

01 · Render

Chromium prints it.
pidief gets it right.

  • Embedded fonts, every glyph in every language
  • Table headers repeat on every page
  • “Page X / Y” footers out of the box

02 · Operations

Then do anything
else to it.

  1. mergepidief merge a.pdf b.pdf c.pdf
  2. splitpidief split report.pdf --span 1
  3. rotatepidief pages rotate scan.pdf --degrees 90
  4. watermarkpidief watermark report.pdf --text DRAFT
  5. compresspidief compress scans.pdf --image-quality 60
  6. PDF/A + encryptpidief pdfa report.pdf && pidief encrypt …
  7. stamppidief stamp report.pdf --text APPROVED

03 · Two doors, one table

The CLI and the API
can’t drift apart.

Both are generated from one operation registry: same names, same parameters, same results.

CLI
html
template
images
merge
split
pages rotate
watermark
stamp
compress
pdfa
encrypt
render
HTTP
200 OK application/pdf

04 · Small and open

No LibreOffice. No AGPL.
No surprises on the invoice.

0MiBDocker image, amd64 + arm64
0MiBidle memory, Chromium warm
0sp50 for a 10-page report
0opsone table, CLI and API
pidief256 MiB
typical Chromium + LibreOffice stack~1.5 GB
  • Go, static binary, no cgo
  • PDFium WebAssembly previews
  • pdfcpu (Apache-2.0)
  • OpenAPI 3.1
  • Prometheus metrics
  • async jobs
  • .NET client on NuGet
  • PDF/A-2b, veraPDF-checked

Ship PDFs today.

Dockerdocker run -p 3000:3000 -e PIDIEF_API_KEYS=change-me ghcr.io/milvasoft/pidief
.NETdotnet add package Pidief.Client
CLIpidief html report.html --page-numbers -o report.pdf