MCP

Paperful provides a Model Context Protocol (MCP) server for giving AI agents direct access to your documents and document intelligence.

Instead of building custom integrations around the REST API, MCP-compatible clients can connect to Paperful and use document tools through a standardized interface.

MCP endpoint

The Paperful MCP server is available at:

https://mcp.paperful.io

Your MCP client connects to this endpoint and authenticates using your Paperful credentials.

What can agents do?

Through the Paperful MCP server, compatible AI clients can work with your document workspace using tools exposed by Paperful.

Depending on the available tools, agents can:

• Find papers
• Retrieve document content
• Inspect structured nodes
• Search documents
• Retrieve relevant context

This makes Paperful useful as a document layer for AI assistants, coding agents, internal tools, and custom agent workflows.

Paperful API vs MCP

Both interfaces provide access to Paperful, but they are intended for different use cases.

Use the API or SDKs when your application controls how Paperful is used.

Use MCP when you want an AI agent to discover and use Paperful capabilities dynamically.

Authentication

The MCP server only exposes resources that the authenticated connection has permission to access.

Keep authentication credentials private and only configure Paperful with MCP clients you trust.