How to Use the Duffel MCP in Pydantic AI

Q: How do I connect the Duffel MCP Server to a Pydantic AI agent?

Use the MCPToolset class in Pydantic AI pointing to your Vinkius HTTP endpoint for Duffel. Pass this toolset into the toolsets argument of your Pydantic AI Agent constructor to auto-discover all 29 Duffel travel tools.

Q: Can I use Pydantic AI to validate Duffel seat maps?

Yes, your Pydantic AI agent can call Duffel's getseatmaps, and the framework will parse the seat coordinates and availability against your defined Pydantic schemas before presenting them to the user.

Q: How do I handle Duffel flight cancellations in Pydantic AI?

Your Pydantic AI agent can invoke Duffel's cancelorder with the order ID. The server processes the cancellation with the carrier and returns the refund status, which is validated by your Pydantic AI Python models.

Build type-safe travel booking agents with Pydantic AI and verify every Duffel flight and hotel offer at runtime.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Duffel MCP to Pydantic AI

Create your Vinkius account to connect Duffel to Pydantic AI and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Duffel with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Type-safe flight search with Pydantic AI

Run Duffel flight searches with complete type safety using Pydantic AI. Your Pydantic AI agent calls Duffel's `create_offer_request` and `list_offers` using this Duffel MCP Server, and Pydantic AI validates the incoming airline data against strict Python models before your code runs. This prevents your Pydantic AI application from crashing due to unexpected Duffel API changes. If an airline returns a weird baggage policy format, the Pydantic AI runtime validation catches it instantly, keeping your Duffel booking pipeline clean.

Strict validation for hotel bookings

Search for accommodations using Duffel's `search_stays` and get verified rates with `fetch_stay_rates` using Pydantic AI. Every room option, price point, and cancellation policy is parsed through Pydantic models to guarantee Duffel data integrity. When your Pydantic AI agent executes Duffel's `book_stay` or `quote_stay`, you know the parameters match the exact schema required. This Duffel MCP integration ensures your Pydantic AI agent never sends malformed booking requests to the hotels.

Secure order creation and payment processing

Avoid costly booking mistakes by validating Duffel orders inside Pydantic AI before execution. Your Pydantic AI agent uses Duffel's `price_offer` to confirm pricing, then calls `create_order` and `create_payment` to secure the tickets. Because Pydantic AI is model-agnostic, you can use any LLM to drive these Duffel tools. If the model tries to pass an invalid currency code to Duffel's `create_payment`, the Pydantic AI validation layer blocks the call before it hits the network.

Setup guide

Set up Duffel 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": {
        "duffel-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

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

result = await agent.run("List recent Duffel 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 Duffel. 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 Duffel 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 Duffel MCP in Pydantic AI

Use the `MCPToolset` class in Pydantic AI pointing to your Vinkius HTTP endpoint for Duffel. Pass this toolset into the `toolsets` argument of your Pydantic AI `Agent` constructor to auto-discover all 29 Duffel travel tools.

The Pydantic AI SDK raises a validation error immediately at runtime when parsing Duffel's response. This prevents your Pydantic AI agent from making decisions based on corrupted or malformed flight or accommodation data.

Yes, your Pydantic AI agent can call Duffel's `get_seat_maps`, and the framework will parse the seat coordinates and availability against your defined Pydantic schemas before presenting them to the user.

Your Pydantic AI agent can invoke Duffel's `cancel_order` with the order ID. The server processes the cancellation with the carrier and returns the refund status, which is validated by your Pydantic AI Python models.

All sensitive data, including passport numbers and billing addresses, is processed inside an isolated V8 container. Vinkius enforces a zero-trust model for this Duffel MCP Server, ensuring travel payloads are never logged or cached on disk during Pydantic AI execution.

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