Set it up
- 1
Turn on Context7 in your AI
In Claude: open Settings → Connectors and add Context7. In ChatGPT: Settings → Apps & Connectors.
- 2
Add it to your app once
Copy the short config below into your app. It just tells your AI where to find Context7.
- 3
Just ask
Tell it what you want in plain words — no special commands.
For example, say
“Look up the current Next.js App Router documentation for server actions and explain it”
Need the exact config? Show it +
Add to your client
{
"mcpServers": {
"context7": {
"command": "npx",
"args": ["-y", "@upstash/context7-mcp"]
}
}
}Paste into your client's MCP configuration file.
Quick skills
Copy one, paste it to your AI, watch it work.
Version-Accurate Code Generation
“Write a Next.js 15 Server Component that fetches data using the latest recommended pattern”
API Reference Lookup
“What's the correct signature and usage for useActionState in React 19?”
Migration Help
“Help me migrate this Prisma 4 schema to Prisma 6 — use the real migration guide”
In the wild
“The Context7 integration alone has saved me from countless "why doesn't this work?" moments where Claude gave me outdated syntax. Worth the five minutes of setup.”
“Context7 does not replace knowing Playwright. It gives the AI better context.”
Good to know
- · Not every library is indexed — less popular ones may fall back to the AI's memory
- · Documentation freshness varies by project — active ones update fast, dormant ones lag
- · You still need to verify generated code — Context7 reduces but doesn't eliminate hallucinations
When to use
- ✓ You want your AI to write code against the actual latest API of a library
- ✓ You keep hitting hallucinated function names or outdated patterns
- ✓ You're using a library that changed significantly between versions
- ✓ You're learning a new library and want your AI to reference real docs
When not to use
- × You're using a proprietary or private library not indexed by Context7
- × You're offline — this needs network access to Context7's index
- × You already have exhaustive docs in your codebase — use a local RAG instead
About
Imagine your AI always having the latest, accurate documentation for any library — so when you ask 'How do I use useSyncExternalStore in React 19?' it reads the real React 19 docs instead of making things up. This server connects your AI to Context7's always-current library documentation index.
The Context7 MCP server resolves library names to current documentation and injects relevant code examples directly into your AI's context. Eliminates hallucinated API calls by providing accurate, version-specific documentation.
Particularly valuable for working with rapidly evolving libraries where LLM training data may be outdated. Supports thousands of popular libraries.
Workflows That Use Context7 MCP
Works With
Also Consider
FAQ
What does Context7 MCP do? +
Imagine your AI always having the latest, accurate documentation for any library — so when you ask 'How do I use useSyncExternalStore in React 19?' it reads the real React 19 docs instead of making things up. This server connects your AI to Context7's always-current library documentation index.
Do I need to know how to code? +
No. Turn it on in your AI's settings and ask in plain English — no terminal, no coding.
When should I use Context7 MCP? +
Reach for it when you need to: You want your AI to write code against the actual latest API of a library; You keep hitting hallucinated function names or outdated patterns; You're using a library that changed significantly between versions; You're learning a new library and want your AI to reference real docs.
When should I avoid Context7 MCP? +
Skip it when: You're using a proprietary or private library not indexed by Context7; You're offline — this needs network access to Context7's index; You already have exhaustive docs in your codebase — use a local RAG instead.
How do I install Context7 MCP? +
In Claude Code, run: claude mcp add context7 -- npx -y @upstash/context7-mcp. Copy-paste configs for Claude Desktop, Cursor, and Windsurf are at the top of this page.
