How to Use the Exa MCP in Pydantic AI

Q: How do I configure the Exa toolset in Pydantic AI?

Import the MCP connector from the pydantic-ai-slimmcp package and pass your Vinkius HTTP URL to it. Then, simply register the toolset by passing it to your Agent constructor in the toolsets list.

Q: Does the Pydantic AI integration support the Exa searchrecent tool?

Yes, your agent can call searchrecent to retrieve validated, time-sensitive web results. The framework ensures that the timestamps and metadata returned match your runtime types exactly.

Q: Can I use Exa to get direct answers within a Pydantic AI agent?

Yes, your agent can invoke the answer tool to get a direct, structured response from the web. This prevents your agent from having to manually extract answers from raw search results.

Inject type-safe, validated web search results into Pydantic AI agents to prevent runtime failures.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Exa MCP to Pydantic AI

Create your Vinkius account to connect Exa 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 Exa with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Type-Safe Web Search for Pydantic AI

The `search` tool delivers validated, structured web results directly to your type-safe agent runtime. By integrating this MCP Server, Pydantic AI validates every search result against strict schemas at runtime, ensuring your agent never processes malformed data. Your agent can safely trigger queries knowing that the resulting URLs, titles, and scores from the MCP Server will perfectly match your defined Pydantic models. If the server returns something unexpected, the framework catches it loudly before it corrupts your state.

Conceptual Research with Strict Validation

The `search_neural` tool executes conceptual queries to find pages that match the actual meaning of the prompt. Traditional keyword search often misses the mark when users ask conceptual questions. To extract the actual text from those matches, your agent can call `search_with_contents`. Pydantic AI will automatically parse and validate the returned page content, giving your agent clean, structured strings to reason over.

Finding and Extracting Similar Sites Safely

The `find_similar` tool allows your agent to discover related web properties based on a seed URL. This is incredibly useful for automated lead generation or brand monitoring. You can also use `find_similar_with_contents` to pull down the text of those related sites in one go. Because Pydantic AI is model-agnostic, you can feed this validated data into any model, whether you are running a local LLM or a commercial API.

Setup guide

Set up Exa 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": {
        "exa-alternative-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

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

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

Import the MCP connector from the `pydantic-ai-slim[mcp]` package and pass your Vinkius HTTP URL to it. Then, simply register the toolset by passing it to your `Agent` constructor in the `toolsets` list.

Yes, your agent can call `search_recent` to retrieve validated, time-sensitive web results. The framework ensures that the timestamps and metadata returned match your runtime types exactly.

Yes, your agent can invoke the `answer` tool to get a direct, structured response from the web. This prevents your agent from having to manually extract answers from raw search results.

The framework will raise a validation or execution error, allowing you to handle the exception gracefully in your Python code rather than letting the agent hallucinate a response.

Every search query and extracted page content payload is handled inside a zero-trust, isolated environment on Vinkius. No data is stored, logged, or shared, keeping your proprietary search strings completely confidential.

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