brightdata-mcp
Bright Data MCP gives AI agents real-time access to public web data through Bright Data's infrastructure. It exposes dozens of tools for web search (Google, Bing and Yandex results as structured data), scraping any URL to Markdown or HTML, extracting clean JSON from major platforms such as Amazon, LinkedIn, Instagram, TikTok, YouTube, X, Reddit, Crunchbase and Zillow, and automating a remote browser session.
It suits teams whose agents keep hitting blocked pages, CAPTCHAs or geo-restrictions, and those who need structured data from large platforms without writing parsers. Every request is routed through Bright Data's unblocking infrastructure, which handles bot detection, CAPTCHA solving and proxy rotation automatically. There is a free tier of 5,000 requests per month with no credit card required, and it can be used as a hosted remote server or run locally.
Getting started: create a Bright Data account and API token, then either add the hosted server URL to your MCP client or run it locally with `npx @brightdata/mcp`, passing the token as `API_TOKEN`.
Features
Why choose it
Trade-offs
Maintenance check
- Last commit
- 2026-11-05
- Commits, last 90 days
- 100+
- Releases, last 90 days
- 13
- Latest release
- v4.6.0 (2026-11-05)
- Open issues
- 37
- Recent issues with at least one comment
- 2/8
Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.
Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.
Most discussed open issues
- #104 Intermittent 502 Bad Gateway errors from MCP endpoint under moderate call volume (no retry guidance)5 comments · 2026-01-19
- #181 Hosted search_engine: organic results contain opaque relative /goto links instead of destination URLs2 comments · 2026-09-10
- #85 I can not connect to my Antigravity2 comments · 2025-11-22
- #2 Password auth with Web Unlocker?2 comments · 2025-04-18
- #136 Proposal: add a Perplexity Search MCP tool as an alternative search source1 comments · 2026-04-29
Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.
Who it fits
- You need to fetch pages that block ordinary requests; the README presents Bright Data's web access and scraping browser as the core of the server.
- You want to try it at no cost first: the README advertises a free tier of 5,000 requests a month in its default Rapid mode.
- You want a hosted endpoint (mcp.brightdata.com/mcp with a token) instead of running the server locally.
- You want a tool that works with no vendor account: API_TOKEN is required, and the README says Pro mode (60+ tools) is not included in the free tier and is billed.
- You only need to drive your own local browser for testing; Playwright MCP or Chrome DevTools MCP need no account.
- You need everything to stay on your machine: requests go through Bright Data's service.
How it differs
Playwright MCP is a local, account-free browser driver. Bright Data MCP is a commercial service wrapper (MIT-licensed client code) whose value is unblocking and structured data extraction from sites that resist scraping, at the price of an API token and usage limits.
Compatibility
Quick start
Use cases
Alternatives
Featured in these lists
Comments
No comments yet. Be the first!