How to Use the PDF.co MCP in Pydantic AI

Q: How do I set up the PDF.co MCP Server in Pydantic AI?

Use the MCPToolset class with your Vinkius HTTP endpoint. Pass the toolset directly into your Agent constructor to expose all 12 document tools to your model.

Q: How does Pydantic AI handle errors from the PDF.co API?

If a tool like pdftojson fails or times out, the framework raises a standard validation or connection exception that you can catch and handle using normal Python try-except blocks.

Q: Does this setup support asynchronous polling for large files?

Yes. You can write an async loop in Python that calls checkjobstatus to monitor long-running extraction tasks without blocking your application's main thread.

Build type-safe document pipelines using the PDF.co MCP Server and Pydantic AI for guaranteed runtime data structures.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers — Included with Plan

Connect PDF.co MCP to Pydantic AI

Create your Vinkius account to connect PDF.co to Pydantic AI — we handle the hosting, security, and runtime updates so you don't have to. No server setup required.

GDPR Included with Plan

Setup PDF.co with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity Vinkius runs on Gemini

Gemini

Grok

Type-safe document parsing with Pydantic AI

Stop guessing if your agent parsed a document correctly. By combining this MCP server with Pydantic AI, you force the outputs of `pdf_to_json` to validate against your strict Pydantic models at runtime. If the API returns unexpected nested structures, the framework throws a clear validation error immediately. This prevents corrupted data from quietly entering your production database.

Validated structural conversions

When converting complex tables, use `pdf_to_xml` or `pdf_to_csv` to extract data. Your agent can then parse this raw text directly into structured Python classes with guaranteed field types. If you need to extract document properties, `extract_pdf_meta` provides structured metadata. The framework ensures that dates, author names, and page counts match your exact schema before processing continues.

Secure, typed file operations

Manage file access securely. Your agent can run `protect_pdf` to restrict file editing, and you can validate the success of the operation using strict boolean return types in your Python code. When stripping security layers, `unprotect_pdf` handles decryption. The framework verifies that the output is a valid, unencrypted PDF URL before passing it to downstream tools.

Setup guide

Set up PDF.co MCP in Pydantic AI

Prerequisites

Python 3.10+ installed
pydantic-ai-slim[fastmcp] package
Active Vinkius subscription with a valid endpoint token

1

Install Pydantic AI with FastMCP
Run pip install "pydantic-ai-slim[fastmcp]". The FastMCP toolset replaces the deprecated MCPServerHTTP class with full protocol support.
2

Configure the FastMCPToolset
Pass a JSON-style config dict to FastMCPToolset with your Vinkius URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. Supports Streamable HTTP, SSE, and Stdio transports.
3

Create and run your agent
Pass the toolset to Agent(toolsets=[toolset]) and call agent.run(). Swap openai:gpt-4o for any supported model — Anthropic, Google, Mistral, or Groq.

agent.py

from pydantic_ai import Agent
from pydantic_ai.toolsets.fastmcp import FastMCPToolset

toolset = FastMCPToolset({
    "mcpServers": {
        "pdfco-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

agent = Agent(
    "openai:gpt-4o",
    toolsets=[toolset],
    system_prompt="You have access to PDF.co tools.",
)

result = await agent.run("List recent PDF.co transactions")
print(result.output)

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by PDF.co. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect PDF.co now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about PDF.co MCP in Pydantic AI

Use the `MCPToolset` class with your Vinkius HTTP endpoint. Pass the toolset directly into your `Agent` constructor to expose all 12 document tools to your model.

If a tool like `pdf_to_json` fails or times out, the framework raises a standard validation or connection exception that you can catch and handle using normal Python try-except blocks.

Yes. You can define a Pydantic model for your target data structure and have the agent parse the raw text returned by `ocr_image` directly into that validated model.

Yes. You can write an async loop in Python that calls `check_job_status` to monitor long-running extraction tasks without blocking your application's main thread.

Your raw PDF files are never stored on the server. They are processed in secure, ephemeral V8 sandboxes via MCP, and all network communication is encrypted via HTTPS to prevent data interception.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript