How to Use the Exa MCP in AutoGen

Q: Can an AutoGen agent use Exa to get quick answers?

Yes, agents can run the answer tool to get a direct, AI-generated summary of web facts. This is perfect for quick checks before starting a deeper debate.

Q: What happens if a search returns too much content for the agent context?

You should use search to get titles and URLs first. The agent can then select the most relevant URL and call getcontents to pull only the text it actually needs.

Let your AutoGen agents debate and verify web facts in real time using semantic search tools.

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 AutoGen

Create your Vinkius account to connect Exa to AutoGen 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 AutoGen

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Multi-agent debate grounded in real web data

The `search_neural` tool provides your AutoGen agents with deep, conceptual web search capabilities during their debate loops. By exposing the MCP server tools to your agents, they can run conceptual queries to resolve disagreements during their conversation. If an agent needs to dig deeper, it can use `get_contents` to pull the full text of a disputed page. This allows the group to analyze the source material directly instead of relying on generic snippets or hallucinations.

Automated competitive analysis using AutoGen and MCP

The `find_similar` tool allows your AutoGen research agents to discover related companies or products based on a seed URL. You can set up a research agent to track competitors and a synthesis agent to write reports. To get the full picture, the research agent can invoke `find_similar_with_contents` to pull both the URLs and the page text in a single turn. This keeps the multi-agent conversation moving fast without waiting for multiple API calls.

Precise domain targeting for specialized agents

The `search_domain` tool restricts your AutoGen developer agents to searching within specific technical documentation sites. This keeps the agent's focus sharp and reduces the noise in the conversation history. If the agent needs exact syntax, it can switch to `search_keyword` to find specific error messages. This structured search approach keeps your conversations highly technical and on-track.

Setup guide

Set up Exa MCP in AutoGen

Prerequisites

Python 3.10+ installed
autogen-ext[mcp] package
Active Vinkius subscription with a valid endpoint token

1

Install AutoGen with MCP
Run pip install "autogen-ext[mcp]" autogen-agentchat. The MCP extension includes mcp_server_tools for stateless tool access.
2

Fetch tools from the MCP
Call mcp_server_tools(SseServerParams(url=...)) with your Vinkius endpoint. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com.
3

Run your agent
Pass the tools to AssistantAgent and call agent.run(). The agent invokes Exa tools and returns structured results.

agent.py

from autogen_ext.tools.mcp import SseServerParams, mcp_server_tools
from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient

server_params = SseServerParams(
    url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
)

tools = await mcp_server_tools(server_params)

agent = AssistantAgent(
    name="Exa_assistant",
    model_client=OpenAIChatCompletionClient(model="gpt-4o"),
    tools=tools,
)

result = await agent.run("List recent Exa data")
print(result.messages[-1].content)

Get your connection token →

Prerequisites

Python 3.10+ installed
autogen-ext[mcp] + autogen-agentchat
Active Vinkius subscription with a valid endpoint token

1

Install dependencies
Same packages as above. McpWorkbench is ideal when your agent needs stateful sessions across multiple tool calls.
2

Use McpWorkbench as context manager
Wrap your agent in async with McpWorkbench(...) to maintain shared state and resources. The workbench manages the full MCP session lifecycle.
3

Run with workbench
Pass workbench=workbench to your agent. State is preserved across multiple tool calls within the same session.

agent.py

from autogen_ext.tools.mcp import McpWorkbench, SseServerParams
from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient

server_params = SseServerParams(
    url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
)

async with McpWorkbench(server_params) as workbench:
    agent = AssistantAgent(
        name="Exa_assistant",
        model_client=OpenAIChatCompletionClient(model="gpt-4o"),
        workbench=workbench,
    )

    result = await agent.run("List recent Exa data")
    print(result.messages[-1].content)

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 AutoGen

The agents negotiate MCP tool usage through their conversational loop. For example, a research agent will call `search_neural` and pass the raw results to a writer agent to draft the final response.

Yes, agents can run the `answer` tool to get a direct, AI-generated summary of web facts. This is perfect for quick checks before starting a deeper debate.

You register the tools with specific agents during setup. Only the designated research agent will have access to `search_keyword` or `get_contents`, keeping other agents focused on analysis.

You should use `search` to get titles and URLs first. The agent can then select the most relevant URL and call `get_contents` to pull only the text it actually needs.

Yes, all web search queries and retrieved page contents are processed on an ephemeral, zero-trust host. No data is stored, and your API keys are never exposed to the agents or saved in logs.

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