PDFops is a hosted HTTP API for the PDF work that keeps getting reinvented: fill AcroForm fields, merge documents, generate invoices. No headless Chromium and no native dependencies, so it runs anywhere a fetch call does, including Vercel, Cloudflare Workers, Lambda, Deno Deploy and Bun. Fills are byte-deterministic, which is what makes them testable in CI: same input, same bytes. There is an MCP server too, so agents can call it over stdio without a key.
Fill AcroForm fields from a JSON map, flatten optional
Merge multiple PDFs in one request
Generate invoices from structured data
Inspect any PDF's form fields, no signup needed
Byte-deterministic output, diffable in CI
No headless browser and no native deps, runs on any edge or serverless runtime
Typed zero-dep SDK, an MCP server and an n8n community node
Filling government and tax forms from application data
Generating invoices and receipts from a serverless function
Bulk-filling contracts or onboarding packets from a CSV
Letting an AI agent fill and merge PDFs over MCP
Merging generated reports into one document for delivery

Built this because every serverless project I touched eventually needed to write a PDF, and the usual answer is ship a headless Chromium, which you cannot do inside a Worker or a small Lambda. PDFops does the fill, merge and invoice cases over plain HTTP instead, with byte-deterministic output so you can diff the result in CI. Happy to answer anything about the determinism guarantees or the MCP server.

Built this because every serverless project I touched eventually needed to write a PDF, and the usual answer is ship a headless Chromium, which you cannot do inside a Worker or a small Lambda. PDFops does the fill, merge and invoice cases over plain HTTP instead, with byte-deterministic output so you can diff the result in CI. Happy to answer anything about the determinism guarantees or the MCP server.
Find your next favorite product or submit your own. Made by @FalakDigital.
Copyright ©2025. All Rights Reserved