Crazyrouter Mcp Server

Name: Crazyrouter Mcp Server
Author: xujfcn

@xujfcn

代码仓库

关于 Crazyrouter Mcp Server

暂无概览

#ai #api #gateway #mcp-server #openai #claude #gemini #image-generation #video-generation #llm

基本信息

分类

其他

传输方式

stdio

发布者

xujfcn

提交者

Jianfeng Xu

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "crazyrouter": {
      "command": "npx",
      "args": [
        "-y",
        "crazyrouter-mcp"
      ],
      "env": {
        "CRAZYROUTER_API_KEY": "<YOUR_API_KEY>"
      }
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Crazyrouter Mcp Server?

An MCP (Model Context Protocol) server that gives AI assistants access to 627+ AI models through a single Crazyrouter API gateway. It supports text, image, video, audio, and music generation, and is designed for users of Claude Desktop, Cursor, VS Code, and other MCP-compatible clients.

How to use Crazyrouter Mcp Server?

Obtain an API key from crazyrouter.com, then clone the repository, run npm install and npm run build. Configure your AI client’s MCP settings with the compiled dist/index.js file and set the CRAZYROUTER_API_KEY environment variable.

Key features of Crazyrouter Mcp Server

627+ models accessible via one API key
Image generation (DALL‑E 3, Midjourney, Flux)
Video generation (Sora 2, Kling V2, Veo 3)
Music and audio generation (Suno, TTS/STT)
Competitive pricing below official provider rates
7 global edge nodes for low‑latency access

Use cases of Crazyrouter Mcp Server

Ask an AI assistant to chat with GPT‑5 or Claude Opus without switching providers
Generate marketing images with DALL‑E 3 or Midjourney on demand
Create short video clips (e.g., ocean waves) using Kling V2 or Sora 2
Compose original music tracks with Suno V4
Transcribe audio or synthesize speech via Whisper‑1 and TTS engines

FAQ from Crazyrouter Mcp Server

How does Crazyrouter Mcp Server compare to using direct provider APIs?

It consolidates 627+ models under one API key, often at rates below official pricing, and supports OpenAI, Anthropic, and Gemini protocols as a drop‑in replacement.

What runtime or dependencies are required?

Node.js 18 or higher, an npm install, and a build step. No additional runtime dependencies beyond the MCP client.

Where does my data go when using this server?

All requests are routed through the Crazyrouter gateway (crazyrouter.com) to the underlying model providers. The server itself does not store data.

What transports or authentication does it support?

It uses the MCP protocol over stdio and authenticates via the CRAZYROUTER_API_KEY environment variable. API endpoints are OpenAI‑compatible HTTP calls behind the scenes.

Are there any known limits?

The README does not list explicit rate limits or model availability caps. Models may change over time; the full list is maintained at crazyrouter.com/models.