Context7, developed by Upstash, is an MCP server designed to enhance AI coding assistants and LLM-powered editors. It delivers up-to-date, version-specific documentation directly into the AI's context during queries. Built with TypeScript/Node.js, it requires Node.js 18 or higher for local use. Context7 is available as npm packages, a VS Code Extension, and Docker containers for enterprise deployments. This server aims to improve the accuracy and relevance of AI-generated code by providing precise documentation for various libraries and frameworks.

What does Context7 connect to?

Context7 functions as a Model Context Protocol (MCP) server, integrating with AI coding assistants and LLM-powered editors or agents. It augments clients such as Cursor, Claude Code (Claude Desktop), Opencode, Windsurf, and Pi Coding Agent. By injecting fresh, version-specific code documentation, Context7 ensures these AI tools have the most current information at query time. It supports HTTP/2/3 transport protocols, including streamable HTTP with plain JSON responses, and a legacy fallback streaming over SSE.

Core features and tools

Context7 exposes specific callable tools to LLMs. These include resolve-library-id, which translates a general library name into a Context7-compatible library ID. Another key tool is query-docs, used for retrieving version-specific documentation for a library based on its Context7 ID. The get-library-docs tool is also mentioned, likely serving a similar function within the VS Code extension and AI SDK. These tools enable AI assistants to access precise and current documentation.

Who is it for?

Context7 targets developers who use AI assistants, library authors, and engineering management. It helps developers building modern applications by ensuring AI assistants provide accurate, version-specific code. Library authors benefit by having their documentation readily available to AI tools. Engineering management can ensure development teams consistently use current and correct information. Skip this if your development workflow does not involve AI coding assistants or LLM-powered editors requiring version-specific documentation.

How to get started

Access is granted through a standard OAuth authorization flow or by configuring an API key. For remote connections, the API key is often passed via HTTP headers, and OAuth tokens are stored locally.

Use cases

• Modern Framework Development: Developers building Next.js 14 apps can ask for middleware setup, receiving version-specific documentation for accurate code generation.
• Cloud Service Integration: Get current configuration examples and API references for platforms like Cloudflare Workers, asking questions such as "Set up Cloudflare Worker with KV storage for caching".
• Library-Specific Implementation: Integrate with libraries like Prisma or Supabase by targeting exact versions and retrieving relevant documentation, potentially using direct library ID syntax.

Context7 by Upstash provides a vital service for AI coding assistants, delivering precise, version-specific documentation. Its support for multiple AI clients and authentication methods makes it adaptable for various development environments. The server's ability to inject current library details directly into LLM contexts significantly enhances code generation accuracy.