← All integrations
Import.io + OpenAI Agents SDK

Web data for the OpenAI Agents SDK

Attach Import.io Web Scraper MCP with MCPServerStreamableHttp and every agent in your workflow can fetch, render and extract from live pages.

Connects via
MCP
Setup
10 min
Available on
10,000 free successful calls, then $0.0002 per call
Get an MCP keyRead the MCP docs
Python
import asyncio, os
from agents import Agent, Runner
from agents.mcp import MCPServerStreamableHttp

async def main():
    async with MCPServerStreamableHttp(
        name="Import.io",
        params={
            "url": "https://mcp.import.io/mcp",
            "headers": {"Authorization": f"Bearer {os.environ['IMPORTIO_MCP_KEY']}"},
        },
    ) as importio:
        agent = Agent(
            name="Price analyst",
            instructions="Fetch live pages with Import.io. Return tables with source URLs.",
            mcp_servers=[importio],
        )
        result = await Runner.run(agent, "Compare prices for AirPods Pro at three US retailers.")
        print(result.final_output)

asyncio.run(main())

How it connects.

Source

The web

Any public page, rendered, through managed proxies

Engine

Import.io

Rendering, proxy routing, captcha handling, extraction

Path

Import.io MCP

https://mcp.import.io/mcp

Destination

OpenAI Agents SDK

Import.io as a Streamable HTTP MCP server for your agents.

What teams do with it.

Analyst agents

A price or market analyst that checks the web before it writes the report.

Compare prices for AirPods Pro at three US retailers.

Handoffs with evidence

A research agent collects pages; a writer agent receives rows and URLs, not guesses.

Collect specs for these five laptops, then hand off to the comparison writer.

Monitoring loops

Run the agent on a schedule and diff today’s extraction against yesterday’s.

Check these competitor pages and report any price change over 5%.

Set it up.

  1. pip install openai-agents.
  2. Open an MCPServerStreamableHttp connection to https://mcp.import.io/mcp with a bearer header.
  3. Pass the server in mcp_servers when you create the agent, then run it.

Questions.

Does the agent list tools on every run?

Tool listing happens when the connection opens. Listing is never billed; only successful tool calls are.

What does the Import.io MCP cost?

Signup includes 10,000 successful calls. After that it is $0.0002 per successful call. Setup requests, tool listings, malformed requests, blocked requests and failed upstream calls are never billed, and you can set a monthly spend limit that stops paid usage when reached.

OAuth or API key?

Use OAuth in interactive clients that support it: the client opens a sign-in, stores the credentials and reuses them. Use an API key for scripts, servers and clients without OAuth, sent as Authorization: Bearer mcp_live_…. Keep keys out of prompts and shared config files.

Which tools does the server expose?

importio_web_fetch, importio_render, importio_extract_data and importio_get_html, page helpers such as importio_page_text, importio_page_links and importio_search_cards, and browser actions for clicks, form input, scrolling, pagination, waits, screenshots and captcha handling. Requests can pick a proxy country and type, with bounded retries when a page blocks.

Related integrations.

All 50 integrations

MCP

OpenAI API

Add Import.io as a remote MCP tool in the Responses API.

Setup 5 minMCP

LangChain & LangGraph

Load Import.io tools with langchain-mcp-adapters into any agent.

Setup 10 minMCP

CrewAI

Give every agent in a crew live web data tools.

Setup 10 minMCP

MCP SDKs

Call Import.io tools directly from the official Python and TypeScript SDKs.

Setup 10 min

Give OpenAI Agents SDK
the live web.

Get an MCP keyRead the MCP docs