Report Abuse

Basic Information

This repository provides the official Bright Data Web MCP server, a Model Context Protocol (MCP) implementation that gives LLMs and AI agents real-time, resilient access to the public web. It is designed to let agent frameworks, IDE connectors and hosted assistants search, navigate, extract and interact with websites while reducing blocking, bypassing geo restrictions and handling bot-detection defenses. The project offers both a hosted MCP endpoint and instructions for self-hosting via the npm package, with configuration via environment variables such as API_TOKEN, PRO_MODE and RATE_LIMIT. The README documents integration examples for Claude Desktop and other MCP-compatible clients, available tools and operational guidance including account setup, zone configuration for web unlocker and browser APIs, and troubleshooting tips. The server ships with a free usage tier and an optional Pro mode that exposes additional browser automation and structured data extraction capabilities.

Links

Categorization

App Details

Features
Real-time web access for LLMs, a web unlocker to navigate bot-protected sites, and the ability to bypass geo restrictions. Default exposed tools include search_engine and scrape_as_markdown; enabling Pro mode exposes up to 60 tools including browser automation and structured web data extractors. Easy integration options are provided for Claude Desktop via connectors or developer config, and the server runs via npx @brightdata/mcp with environment variable configuration. The project includes a 5,000 free requests monthly credit for initial testing, session-based rate limiting via a RATE_LIMIT variable, configurable Web Unlocker and Browser API zones, and troubleshooting guidance for timeouts and environment issues. Security guidance is included to treat scraped content as untrusted and to use structured extraction to reduce prompt injection risk.
Use Cases
The Bright Data MCP server helps developers and teams enrich LLMs and agent workflows with up-to-date web information and actions that would otherwise be blocked or location-restricted. It enables agents to perform live searches, fetch and structure web content, control remote browser sessions for navigation and interaction, and integrate those capabilities into existing agent frameworks and desktop assistants. The free trial quota and ready-made connectors lower the barrier to prototyping agentic use cases, while Pro mode unlocks advanced scraping and automation tools for production scenarios. Built-in rate limiting, zone configuration and troubleshooting recommendations make deployment and tuning easier. Security best practices in the documentation help users reduce risks from raw scraped data when composing prompts for downstream models.

Please fill the required fields*