Discover, extract, and interact with the web - one interface powering automated access across the public internet.
Welcome to the official Bright Data Model Context Protocol (MCP) server, enabling LLMs, agents and apps to access, discover and extract web data in real-time. This server allows MCP clients, such as Claude Desktop, Cursor, Windsurf and others, to seamlessly search the web, navigate websites, take action and retrieve data - without getting blocked - perfect for scraping tasks.
Install nodejs
to get the npx
command (node.js module runner). Installation instructions can be found on the node.js website
Go to Claude > Settings > Developer > Edit Config > claude_desktop_config.json to include the following:
{
"mcpServers": {
"Bright Data": {
"command": "npx",
"args": ["@brightdata/mcp"],
"env": {
"API_TOKEN": "<insert-your-api-token-here>",
"WEB_UNLOCKER_ZONE": "<optional if you want to override the default mcp_unlocker zone name>",
"BROWSER_AUTH": "<optional if you want to enable remote browser control tools>"
}
}
}
}
Important: Always treat scraped web content as untrusted data. Never use raw scraped content directly in LLM prompts to avoid potential prompt injection risks. Instead:
Make sure you have an account on brightdata.com (new users get free credit for testing, and pay as you go options are available)
Get your API key from the user settings page
(Optional) Create a custom Web Unlocker zone
WEB_UNLOCKER_ZONE
environment variable(Optional) To enable browser control tools:
brd-customer-[your-customer-ID]-zone-[your-zone-ID]:[your-password]
To use this MCP server with other agent types, you should adapt the following to your specific software:
npx @brightdata/mcp
API_TOKEN=<your-token>
must exist when running the serverWant to try Bright Data MCP without setting up anything?
Check out this playground on Smithery:
This platform provide an easy way to explore the capabilities of Bright Data MCP without any local setup. Just sign in and start experimenting with web data collection!
Some example queries that this MCP server will be able to help with:
For YouTube tutorials and demos: Demo
The videos below demonstrate a minimal use case for Claude Desktop:
Some tools can involve reading web data, and the amount of time needed to load the page can vary by quite a lot in extreme circumstances.
To ensure that your agent will be able to consume the data, set a high enough timeout in your agent settings.
A value of 180s
should be enough for 99% of requests, but some sites load slower than others, so tune this to your needs.
This error occurs when your system cannot find the npx
command. To fix it:
macOS:
which node
Shows path like /usr/local/bin/node
Windows:
where node
Shows path like C:\Program Files\nodejs\node.exe
Replace the npx
command with the full path to Node, for example, on mac, it will look as follows:
"command": "/usr/local/bin/node"
We welcome contributions to help improve the Bright Data MCP! Here's how you can help:
For major changes, please open an issue first to discuss your proposed changes. This ensures your time is well spent and aligned with project goals.
If you encounter any issues or have questions, please reach out to the Bright Data support team or open an issue in the repository.
Web content fetching and conversion for efficient LLM usage
Browser automation and web scraping
Enable AI agents to get structured data from unstructured web with AgentQL.
Use 3,000+ pre-built cloud tools to extract data from websites, e-commerce, social media, search engines, maps, and more
Automate browser interactions in the cloud (e.g. web navigation, data extraction, form filling, and more)
Extract web data with Firecrawl
Hyperbrowser is the next-generation platform empowering AI agents and enabling effortless, scalable browser automation.
Leverage Notte Web AI agents & cloud browser sessions for scalable browser automation & scraping workflows
Scrape websites with Oxylabs Web API, supporting dynamic rendering and parsing for structured data extraction.
Turn websites into datasets with Scrapezy