PDF Blocks
PricingSupport
Start Free
Go to Page

Libraries and integrations

There is no official SDK yet — raw HTTP is the supported path. Where to find the Power Automate and Zapier connectors, generate a typed client, and MCP on the roadmap.

There is no official SDK yet. PDF Blocks is a clean HTTP API — one PDF in, one PDF out — so raw HTTP is the supported, first-class path, and it works from every language without a dependency to install. This page sets that expectation and points each kind of reader to the right integration surface.

Raw HTTP is first-class

Every action is a single multipart/form-data POST that returns a PDF. The Quickstart gets you a working call in a few minutes, and every action reference page ships complete, runnable snippets in cURL, Python, Ruby, Node.js, and C#. Paste one, swap in your key, and run.

If you want a typed client, generate one from the machine-readable contract — see OpenAPI specification for the codegen tools and commands. That is the recommended way to get language-native models and request methods until an official SDK exists.

Where to go next

Choosing a surface

  • Writing backend code or a script? Call the API directly over HTTP, or generate a typed client from the OpenAPI spec.
  • Building a no-code or low-code automation? Use the Power Automate or Zapier connector.

MCP — on the roadmap

A Model Context Protocol server, which would let AI agents call PDF Blocks actions as tools, is on the roadmap and not yet available. There is nothing to install today. Until it ships, agents should call the HTTP API directly.