Share
X has launched an official hosted Model Context Protocol (MCP) server that lets AI agents connect to the X API without self-hosting infrastructure, giving developers a faster way to build applications that rely on live conversations, trends, and breaking news. The release also introduces a hosted documentation server, allowing compatible AI tools to search X’s developer documentation without leaving their workflow.
The launch comes as AI developers increasingly look for ways to connect language models with live data instead of relying solely on static training data sets. AI agents increasingly need real-time information to complete tasks such as monitoring news, tracking market sentiment, or responding to changing events.
𝕏 just launched hosted MCP for the 𝕏 API
That means AI tools like Grok Build, Cursor, Claude Desktop, VS Code, and other MCP-compatible clients can now connect to 𝕏 without developers needing to host their own MCP server
Agents can use 𝕏 to:
• Search posts
• Track trends… https://t.co/MfyToe0qgl pic.twitter.com/ltHKCjzBDa— X Freeze (@XFreeze) June 30, 2026
The hosted MCP server supports MCP-compatible clients, including Grok Build, Cursor, Claude Desktop, Visual Studio Code, and other AI tools. Instead of deploying and maintaining their own MCP infrastructure, developers can connect through X’s hosted service after authenticating with their X Developer account. The platform also launched a separate Docs MCP server that enables AI assistants to search and retrieve X API documentation during development.
The hosted service exposes X API capabilities covering post search, user lookups, trends, news, bookmarks, timelines, mentions, and Articles. X routes authentication through its open-source Xurl bridge, which manages “OAuth” login and refreshes access tokens automatically, handling authentication without requiring developers to manage OAuth tokens manually.
The launch reflects the growing role of Model Context Protocol as AI agents move beyond static knowledge. The open standard allows AI models to connect with external tools and data sources, enabling agents to retrieve live information instead of relying only on knowledge captured during training.
X’s hosted approach lowers the technical barrier to building AI agents that rely on live X data. Developers still need an X Developer application and the required API credentials, meaning the launch streamlines deployment rather than eliminating authentication or platform access requirements.
The hosted MCP service also expands X’s role in AI workflows built around current events and public conversations. As more AI applications shift toward real-time decision-making, access to continuously updated social data has become increasingly valuable. With hosted MCP services now available for both the X API and its developer documentation, developers can connect AI agents to live platform data through X’s officially supported MCP infrastructure.
Share
preferred on
