Qwen_Max

Qwen_Max

A Model Context Protocol (MCP) server implementation for the Qwen models.

23
Stars
8
Forks
0
Releases

Overview

Qwen Max MCP Server is a Node.js/TypeScript MCP server that enables access to the Qwen model family (Qwen Max, Qwen Plus, and Qwen Turbo) via the MCP protocol, with direct Claude Desktop integration. Prerequisites include Node.js v18+, npm, Claude Desktop, and a Dashscope API key. It can be installed automatically via Smithery or manually from GitHub, and the default model is Qwen-Max, with options to switch to Qwen-Plus or Qwen-Turbo by editing src/index.ts. The server exposes distinct context windows and token budgets per model: Qwen-Max (32,768 context window; 30,720 max input; 8,192 max output; pricing for input/output and a 1M free quota), Qwen-Plus (131,072 context window; 129,024 max input; 8,192 max output; separate pricing and 1M free quota), and Qwen-Turbo (1,000,000 token context; 1,000,000 max input; 8,192 max output; pricing and 1M free quota). Configuration relies on a .env file with DASHSCOPE_API_KEY and a Claude Desktop config mapping the built server (index.js). Development commands, API usage example, and error handling emphasis are provided, all aimed at a robust MCP-compliant gateway to Qwen models.

Details

Owner
66julienmartin
Language
JavaScript
License
MIT License
Updated
2025-12-07

Features

Text generation with Qwen models

Provides text generation using the Qwen model family (Max, Plus, Turbo).

Configurable parameters (max_tokens, temperature)

Expose adjustable generation controls to tailor outputs (max_tokens and temperature).

Error handling

Detailed error messages for API authentication, invalid parameters, rate limits, network issues, token limits, and model availability.

MCP protocol support

Implements MCP protocol compatibility to facilitate communication with Claude Desktop and MCP clients.

Claude Desktop integration

Includes seamless integration with Claude Desktop for MCP-based workflows.

Support for all Qwen commercial models (Max, Plus, Turbo)

Supports the full Qwen commercial model lineup with configurable model selection.

Extensive token context windows

Supports large context windows per model: Qwen-Max (32,768 tokens), Qwen-Plus (131,072 tokens), Qwen-Turbo (1,000,000 tokens).

Audience

MCP developersBuild, deploy, and maintain an MCP server for Qwen models (Max, Plus, Turbo) and enable Claude Desktop integration.
Claude Desktop integratorsConfigure Claude Desktop to connect to the Qwen MCP server via the provided mcpServers configuration and environment variable (DASHSCOPE_API_KEY).

Tags

MCPQwenQwen-MaxQwen-PlusQwen-TurboNode.jsTypeScriptClaude DesktopMCP serverNLPLanguage Model