# Context7 — Live Library Docs for LLMs, 51K Stars

> MCP server fetching up-to-date, version-specific library documentation into LLM context. Eliminates hallucinated APIs across 30+ AI coding tools.

## Install

Merge the JSON below into your `.mcp.json`:

## Quick Use

Run the automated setup:
```bash
npx ctx7 setup
```

Or add to your `.mcp.json` manually:
```json
{
  "mcpServers": {
    "context7": {
      "url": "https://mcp.context7.com/mcp"
    }
  }
}
```

Restart your AI tool and ask: "Using context7, look up the Next.js 14 App Router docs."

---

## Intro

Context7 is an MCP server that fetches live, version-specific library documentation directly into your LLM's context window, with 51,700+ GitHub stars. It eliminates hallucinated or outdated APIs by pulling real-time docs from library sources. Works with 30+ coding clients including Claude Code, Cursor, GitHub Copilot, and Windsurf.

Best for: developers tired of AI hallucinating outdated API methods. Works with: Claude Code, Cursor, Copilot, Windsurf, 30+ tools. Setup time: under 1 minute.

---

## Context7 — Real-Time Docs for AI

### The Problem

LLMs frequently hallucinate API methods or reference outdated library versions. This leads to broken code and debugging time wasted on non-existent functions.

### The Solution

Context7 fetches current documentation from the actual library source and injects it into your AI's context window, ensuring every code suggestion uses real, up-to-date APIs.

### Two Operation Modes

1. **CLI + Skills** — Use `ctx7` commands directly
2. **MCP Server** — Native integration into your coding agent

### MCP Tools

| Tool | Description |
|------|-------------|
| `resolve-library-id` | Match library names to Context7 identifiers |
| `query-docs` | Fetch relevant documentation using a library ID |

### CLI Commands

```bash
# Find a library
ctx7 library nextjs "app router"

# Get specific docs
ctx7 docs nextjs-14 "server components"
```

### Key Benefits

- **Version-specific**: Request docs for Next.js 14, React 19, etc.
- **Real-time**: Always fetches current documentation
- **Token-efficient**: Only retrieves relevant sections, not entire docs
- **30+ integrations**: Works with practically any AI coding tool

### FAQ

**Q: What is Context7?**
A: Context7 is an MCP server that injects up-to-date, version-specific library documentation into AI coding tools, preventing hallucinated API suggestions.

**Q: Is Context7 free?**
A: Yes, Context7 is open source under the MIT license.

**Q: How do I add Context7 to Claude Code?**
A: Run `npx ctx7 setup` or add `{"url": "https://mcp.context7.com/mcp"}` to your `.mcp.json` under mcpServers.

---

## Source & Thanks

> Created by [Upstash](https://github.com/upstash). Licensed under MIT.
>
> [context7](https://github.com/upstash/context7) — ⭐ 51,700+

Thank you for solving the hallucinated API problem.

---

<!-- ZH -->

## 快速使用

运行自动设置:
```bash
npx ctx7 setup
```

或手动添加到 `.mcp.json`:
```json
{
  "mcpServers": {
    "context7": {
      "url": "https://mcp.context7.com/mcp"
    }
  }
}
```

---

## 简介

Context7 是一个 MCP 服务器，可以将实时的、版本特定的库文档直接注入 LLM 的上下文窗口，拥有 51,700+ GitHub stars。它从库源头拉取真实文档，消除 AI 编码工具中的 API 幻觉问题。支持 Claude Code、Cursor、Copilot 等 30+ 编码工具。

---

## Context7 — AI 的实时文档

### 问题

LLM 经常幻觉出不存在的 API 方法或引用过时的库版本。

### 解决方案

Context7 从实际的库源头获取当前文档，并注入到 AI 的上下文窗口中，确保每个代码建议都使用真实、最新的 API。

---

## 来源与感谢

> Created by [Upstash](https://github.com/upstash). Licensed under MIT.
>
> [context7](https://github.com/upstash/context7) — ⭐ 51,700+

---
Source: https://tokrepo.com/en/workflows/d95ee526-658a-4dc4-8f33-d0b6395f7f5c
Author: TokRepo精选