Open-source web data platform that turns websites into structured APIs for scraping, crawling, search, monitoring and AI extraction.
18.2K+61 this week1.6KTypeScriptAGPL-3.0Updated 2 days ago
brightdata/brightdata-mcp
An MCP server giving AI agents web search, page scraping, structured data extraction, and remote browser automation through Bright Data.
- Stars
- 2,712
- Forks
- 337
- Open issues
- 40
- Last commit
- 2 days ago
What is it?
Bright Data MCP is a Model Context Protocol server that lets AI agents, coding agents and chat assistants read live public web data. It exposes 32 tools for searching Google, Bing and Yandex, fetching any URL as Markdown or HTML, and driving a remote browser session with clicks, typing, screenshots and page reads.
For sites that require structured output, the server provides access to more than 1,200 pre-built scrapers covering platforms such as Amazon, LinkedIn, Instagram, TikTok, YouTube, Reddit, npm and PyPI. Each request is routed through Bright Data's unblocking infrastructure, which handles bot detection, CAPTCHAs and proxy rotation. The server also collects ChatGPT and Perplexity answers as structured data.
It can be used as a hosted remote server by adding a URL with an API token to an MCP client, or run locally with npx. Tool groups and individual tool names can be selected through configuration to limit the tools loaded into an agent's context. A free tier of 5,000 requests per month is offered, with pay-as-you-go pricing beyond that.
Key features
- 01Searches Google, Bing and Yandex and returns results as structured data or Markdown.
- 02Fetches any URL as Markdown or HTML with bot detection and CAPTCHA handling built in.
- 03Provides access to 1,200+ pre-built scrapers for sites like Amazon, LinkedIn and TikTok.
- 04Offers 14 remote browser automation tools for navigation, forms, screenshots and page reads.
- 05Collects ChatGPT and Perplexity responses as structured output.
- 06Supports tool groups and individual tool selection to limit context size.
Use cases
- Giving a coding or research agent live web search when built-in fetch tools get blocked.
- Comparing product prices, ratings and stock across retailers like Amazon and Walmart.
- Pulling npm or PyPI package metadata and GitHub files on demand for a coding agent.
- Monitoring how ChatGPT and Perplexity describe a brand or which sources they cite.
Quick start
- 1
Get your API token from your Bright Data account settings
- 2
Add the hosted server URL with your token to your MCP client (e.g. Claude Code)
claude mcp add --transport http brightdata "https://mcp.brightdata.com/mcp?token=YOUR_API_TOKEN" - 3
Or run the server locally with npx using your API token in the client config
npx @brightdata/mcp
Things to know
- Requires a Bright Data API token, and requests stop once the free tier and any deposited funds run out.
- Browser navigation and search-scraping usage is billed beyond the 5,000 monthly free requests.
In collections
Topics
#mcp#web-scraping#web-search#browser-automation#llm#ai-agents#data-extraction#ai-integrations#anti-bot-detection#data-collection
FAQ
- What is brightdata-mcp?
- An MCP server giving AI agents web search, page scraping, structured data extraction, and remote browser automation through Bright Data. Bright Data MCP is a Model Context Protocol server that lets AI agents, coding agents and chat assistants read live public web data. It exposes 32 tools for searching Google, Bing and Yandex, fetching any URL as Markdown or HTML, and driving a remote browser session with clicks, typing, screenshots and page reads.
- Is brightdata-mcp free to use? What license does it have?
- brightdata-mcp is open source under the MIT license. Check the license terms before using it in commercial products.
- Is brightdata-mcp still maintained?
- Actively maintained. The last commit to brightdata-mcp was on Nov 5, 2026.
- What can I use brightdata-mcp for?
- Giving a coding or research agent live web search when built-in fetch tools get blocked. Comparing product prices, ratings and stock across retailers like Amazon and Walmart. Pulling npm or PyPI package metadata and GitHub files on demand for a coding agent. Monitoring how ChatGPT and Perplexity describe a brand or which sources they cite.
- How do I install brightdata-mcp?
- Get your API token from your Bright Data account settings Add the hosted server URL with your token to your MCP client (e.g. Claude Code): claude mcp add --transport http brightdata "https://mcp.brightdata.com/mcp?token=YOUR_API_TOKEN" Or run the server locally with npx using your API token in the client config: npx @brightdata/mcp
- Where can I get brightdata-mcp?
- The source code and installation instructions are on GitHub: https://github.com/brightdata/brightdata-mcp
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Rust web crawler and extraction engine that turns websites, PDFs and Office files into clean Markdown and structured data for AI agents and RAG pipelines.
245+70 this week32RustAGPL-3.0Well maintainedUpdated 19 hours ago
TypeScript, Python and Go SDK for browser agents that drives Chrome with natural-language act, observe and extract calls alongside Playwright-style APIs.
26K+87 this week1.8KTypeScriptMITUpdated 14 hours ago
An API for AI agents to search, scrape, and interact with web pages, returning clean markdown, JSON, and screenshots.
199.2K+2.2K this week10.5KTypeScriptAGPL-3.0Updated 15 hours ago
Official MCP server that gives AI agents and MCP clients web search, scraping, crawling, browser interaction and Markdown or structured-data extraction via Firecrawl.
7.8K+41 this week929JavaScriptMITUpdated 14 hours ago
An MCP server that lets AI agents run any Apify Store scraper or automation Actor to extract data from social media, search engines, maps, and websites.
15.1K+1.1K this week436TypeScriptMITUpdated 20 hours ago