TLDR: Bright Data has launched a free version of The Web MCP (Model Context Protocol), an infrastructure layer designed to provide AI agents with seamless, unblocked access to the live web. This solution aims to overcome a critical limitation in agentic AI by enabling real-time interaction with websites, bypassing common barriers like CAPTCHAs and geo-restrictions, and delivering structured data for immediate action.
NEW YORK – August 12, 2025 – Bright Data, a prominent web data infrastructure company, today announced the public launch of a free version of The Web MCP (Model Context Protocol). This innovative infrastructure layer is specifically engineered to enable AI agents to interact with the live internet without encountering common blocking mechanisms, addressing a significant hurdle in the advancement of agentic AI.
The Web MCP serves as a crucial interface, allowing autonomous AI agents to seamlessly access the open web. It facilitates real-time interaction across various websites, effectively bypassing challenges such as CAPTCHAs, geo-restrictions, and automated traffic filters. A key benefit of the system is its ability to return structured, high-confidence results that AI agents can immediately act upon, enhancing their operational efficiency and reliability.
Or Lenchner, CEO of Bright Data, emphasized the necessity of this development, stating, “The internet wasn’t designed for agents, but The Web MCP is. Agents often fail simply because they lack proper access.” He added, “By making The Web MCP freely available to the public, we’re opening the gateway to the internet that serves both humans and machines alike.”
The platform has already demonstrated significant utility, powering over 100 million daily agent interactions. Its public release follows a successful three-month private beta program that involved 15,000 developers. Bright Data is now offering a free tier that includes 5,000 requests per month, with automatic renewal, making it accessible to a broader developer community.
Compatibility is a cornerstone of The Web MCP, as it is fully compatible with all major Large Language Models (LLMs) and Integrated Development Environments (IDEs), including support for locally hosted models, Server-Sent Events (SSE), and Streamable HTTP. Furthermore, it integrates out-of-the-box with popular AI frameworks such as LangChain, CrewAI, and LlamaIndex.
The Web MCP product offers two distinct access modes to cater to varying needs:
Rapid Mode: This is the default mode, optimized for token efficiency and speed, making it ideal for search and data extraction tasks.
Advanced Mode: Available on paid plans, this mode provides remote browser capabilities, advanced structured content retrieval, and site automation, suitable for more complex, multi-step operations.
Also Read:
- Operant AI Unveils MCP Gateway to Enhance Security and Management for Enterprise AI Deployments
- Kiwi.com Unveils Agentic AI System for Direct Flight Bookings
Key features across both modes include real-time web access for instant data retrieval, a robust web unlocker to bypass various defenses, browser control for automating interactions with dynamic sites, and the output of data in clean JSON format for easy integration. With The Web MCP, AI agents can perform a wide array of tasks, such as booking travel, pulling real-time pricing data, analyzing competitor activity, monitoring regulatory content, aggregating healthcare research, and tracking social trends.
