MCP-Lingvanex-Translate
공식Lingvanex MCP 서버 클라우드 API – Lingvanex 클라우드 API를 사용하여 빠르고 정확한 텍스트 번역을 제공하는 모델 컨텍스트 프로토콜 서버입니다. 100개 이상의 언어를 지원하며 Claude Desktop 및 기타 MCP 호환 클라이언트와 원활하게 통합됩니다.
Lingvanex Translate MCP(으)로 무엇을 할 수 있나요?
- 언어 간 텍스트 번역 — AI 어시스턴트에게
translate_text를 사용하여 텍스트, 소스 언어 및 대상 언어 코드를 지정해 문구를 번역하도록 요청하세요. - Claude Desktop에 번역 기능 통합 —
claude_desktop_config.json에서 서버를 구성하여 Claude가 대화 중에translate_text를 직접 호출할 수 있도록 하세요. - HTTP를 통한 서버 테스트 실행 — HTTP 모드로 서버를 시작하고
curl또는 MCP Inspector를 사용하여/ping엔드포인트와 번역 흐름을 확인하세요.
문서
MCP Prototype – Translate Server
This project implements an MCP (Model Context Protocol) server for text translation. The server supports two transports:
- stdio – for integration with Claude Desktop
- http (streamable) – for testing and working via HTTP + SSE
⚙️ Requirements
- Node.js >= 18
- Yarn or npm
- Installed Claude Desktop (for stdio integration)
- Lingvanex Translator account for text translation
🔑 Lingvanex Translator Setup
To use the Lingvanex Translator you'll need a Lingvanex account.
-
If you don't have one, sign up for free
-
Go to the Cloud API tab: Cloud API
-
Fill out the Billing Address data
-
Click Continue to payment
- To get a free trial, it is not necessary to add your payment card
-
Your API key will be generated and visible in the Cloud API tab: API key
Now you are ready to start using the translation API. Below is a video tutorial of the overall process (if available on Lingvanex site).
🚀 Installation & Build
# Clone the repository
git clone https://github.com/lingvanex-mt/MCP-Lingvanex-Translate.git
cd mcp-prototype
Install dependencies
yarn install
🔌 Run in stdio mode (Claude Desktop)
stdio mode is used by Claude Desktop to connect to local MCP servers.
Set environment variable:
TRANSPORT=stdio
Start the server:
yarn build
yarn start
Expected output:
MCP stdio transport running
Translate MCP Server ready
🌐 Run in HTTP mode (streamable)
http mode runs a local HTTP server with HTTP transport.
Useful for browser testing or with curl.
Set environment variables:
TRANSPORT=http
HTTP_PORT=3000
Start the server:
yarn build
yarn start
Test the server:
curl http://127.0.0.1:3000/ping
Expected response:
{ "status": "ok", "transport": "http" }
Use MCP Inspector for debugging:
npx @modelcontextprotocol/inspector
In the MCP Inspector UI, select Transport Type - Streamable HTTP; URL - http://localhost:3000/mcp. Click Connect.
🖥️ Integration with Claude Desktop
Claude Desktop discovers local MCP servers via config file:
- Windows:
%APPDATA%\Claude\claude_desktop_config.json - macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
Example config (Windows)
Open (or create) claude_desktop_config.json and add:
{
"mcpServers": {
"translate": {
"command": "node",
"args": [
"C:\\Users\\path\\to\\project\\dist\\index.js"
]
}
}
}
⚠️ Make sure to update the path to your local
dist/index.jsafter build!
✅ How to verify
- Launch Claude Desktop.
- Enter a request like:
"Use the MCP tool
translate_textto translate 'Hello world' into Russian." - If everything is configured correctly, Claude will call your MCP server and return the translation.
📌 Available Tools
translate_text
Translate text from one language into another.
Arguments:
text– the text to translatesourceLang– source language code (e.g."en")targetLang– target language code (e.g."ru")
Example request:
{
"tool": "translate_text",
"args": {
"text": "Good morning",
"sourceLang": "en",
"targetLang": "fr"
}
}
Example response:
{
"content": [
{ "type": "text", "text": "Bonjour" }
]
}