Skip to content

MCP server overview

What the PDF As You Go MCP server is, the endpoint it lives at, and the tools it exposes to your agent.

The MCP server is the primary surface of PDF As You Go. It implements the Model Context Protocol, so any MCP-capable agent — Claude, ChatGPT, or a custom LLM app — can discover and call the PDF operations as tools.

https://mcp.pdfasyougo.com

The server speaks MCP over a remote (Streamable HTTP) transport. You connect by pointing your MCP client at this URL and giving it a wallet to pay with. There is no API key. Get the endpoint confirmed for your account via Request preview access.

Each of the 12 operations is an MCP tool:

Tool Operation
merge_pdfs Merge documents
add_text_watermark Add text watermark
add_image_watermark Add image watermark
add_password Add password
remove_password Remove password
add_restrictions Add restrictions
remove_restrictions Remove restrictions
extract_pages Extract pages
remove_pages Remove pages
rotate_pages Rotate pages
reverse_pages Reverse pages
remove_signatures Remove signatures

Each tool’s input schema mirrors the parameters documented on its operation page, and each call costs a flat $0.05. See the Tools reference for the parameter summary.

  1. Your agent decides to call a tool — say merge_pdfs — with arguments.
  2. The server responds that payment is required.
  3. Your x402-capable client signs and pays, transparently.
  4. The operation runs and the tool returns a short-lived download URL.

From the model’s perspective it’s a single tool call that returns a result; the payment happens underneath. See How it works for the end-to-end picture.

MCP is becoming the standard way LLMs call external tools. Exposing PDF operations as MCP tools means an agent can pick them up with no glue code — discover the tool, read its schema, call it. Pairing that with x402 means the agent can also pay for the call on its own, which is the whole point of PDF As You Go.