PDF. Solved.

The complete PDF engine: rendering, processing, editing and conversion, in one container on your own servers.

Render pages faithfully, merge, split and compress, edit and redact text, fill forms, and convert to Word, Markdown or JSON through one plain HTTP API. No calls home, no per-page fees.

  • In production on zonepdf.com
  • Evaluate without signing up
  • linux/amd64 and arm64
# evaluate: no key, no sign-up
$ docker run --rm -p 8080:8080 --read-only zonepdf/engine

# a smaller PDF, a Word file, page images
$ curl -F file=@report.pdf -F quality=balanced \
    -o smaller.pdf localhost:8080/v1/compress
$ curl -F file=@report.pdf -o report.docx \
    localhost:8080/v1/to-docx
$ curl -F file=@report.pdf -F dpi=150 -o pages.zip \
    localhost:8080/v1/to-image

Documents never leave your network

Why teams self-host it

Nothing leaves your network

The engine makes no network calls. Its license is checked offline, so it runs the same on an air-gapped network.

One binary to patch

No third-party software underneath. One self-contained binary, fuzz-tested, running as an unprivileged user on a read-only filesystem.

Licensing legal can sign off

No GPL or AGPL code in the engine, fonts licensed for commercial bundling, and a flat yearly price with no per-page metering.

One endpoint per job

Send the document as multipart form data or as the raw request body. A result comes back as the file, several come back as a ZIP, and errors come back as JSON.

Render and compress

  • /v1/to-imagePages to PNG, JPEG, TIFF, BMP or GIF
  • /v1/from-imageImages to a PDF, one page each
  • /v1/compressMake files smaller, from lossless to maximum
  • /v1/grayscaleMake every color gray, nothing else changed

Pages

  • /v1/mergeMerge documents in the order uploaded
  • /v1/splitExtract pages, or split into several files
  • /v1/rotateTurn pages
  • /v1/pagesRebuild a document from a page plan
  • /v1/numberAdd page numbers
  • /v1/watermarkLay text or a picture on pages

Edit and redact

  • /v1/edit/findFind text
  • /v1/edit/replaceReplace every match
  • /v1/edit/redactRemove every match from the file, verified
  • /v1/edit/applyApply a list of edits

Forms and annotations

  • /v1/form/fillFill form fields
  • /v1/form/flattenPaint fields into the pages
  • /v1/annot/applyAdd, update or delete annotations
  • /v1/annot/flattenPaint annotations into the pages

Convert

  • /v1/to-docxPDF to Word
  • /v1/to-xlsxPDF to Excelbeta
  • /v1/to-pptxPDF to PowerPointbeta
  • /v1/to-markdownPDF to Markdown
  • /v1/to-jsonStructured JSON, figures on request
  • /v1/to-htmlPDF to HTML
  • /v1/to-textText in reading order

Inspect, repair and secure

  • /v1/infoPages, geometry, permissions and outline
  • /v1/repairRebuild a damaged file
  • /v1/signaturesSignature fields and what they hold
  • /v1/encryptEncrypt, AES-256 by default
  • /v1/decryptRemove the encryption

Every instance serves its full OpenAPI reference at /v1/openapi.yaml, and answers your probes at /health and /readyz. Without a key, each document is cut to its first 5 pages and PDF and image results carry a small evaluation mark.

Priced by capacity, not by the page

Every plan includes unlimited documents and pages, every endpoint, updates for the term, and test and staging servers at no extra cost. Billed yearly.

Starter

$1,500

per year

  • Up to 4 documents at once per server
  • One production server
  • Email support
Get Starter

Enterprise

from $20,000

per year

  • No limit on documents at once
  • Any number of production servers
  • Support SLA, security review, invoice and PO
Contact us

Pick a plan and we reply with payment details and your key, which goes in PDFENGINE_LICENSE or a mounted file; a renewal is picked up without a restart. Shipping the engine inside your own product? OEM licenses are quoted: tell us how you'd deploy it.

Frequently asked questions

Is it the same engine zonepdf.com runs on?

Yes. Every merge, split, compression, page render and PDF to Word conversion on zonepdf.com goes through it.

Does it create PDFs from HTML, Word or Excel files?

No. The engine works on PDFs that already exist: it reads, changes, renders and converts them. To make PDFs from HTML or Office files, use a tool built for that and pass the result to the engine.

Does it send our documents or usage anywhere?

No. The engine makes no network calls. Its license key is checked offline against a signature built into the binary, so it runs the same on an air-gapped network.

Is the source code available?

No. ZonePDF Engine is proprietary and ships as a Docker image. A license covers running it, inside your own product too under an OEM license, but never the source.

What does evaluation mode leave out?

Each document is cut to its first 5 pages, and every page of a PDF or image result carries a small red evaluation mark. Every endpoint and option works, with no sign-up and no time limit.

How are plans measured?

By capacity, never by volume. Your key sets how many documents a server works on at once, and your plan covers a number of production servers. Documents and pages are never counted.

What happens when a license runs out?

The engine keeps working for 14 days, with a warning header on every response and a warning in its log, then drops to evaluation mode. A renewed key file is picked up without a restart.

Can it read scanned documents?

It reads the text a PDF carries. There is no OCR in the engine, so a scanned page without a text layer gives no text.

Run it on your servers today

Pull the image and evaluate without signing up. When you need whole documents, a license key is one email away.

Get a license key