DeepL

You can run a dedicated MCP Server that accesses DeepL translation capabilities, making it easy to translate text, adjust formality, and query language options from your existing MCP client setup. This server bridges your client with the DeepL API and manages translation tasks efficiently within your workflow.

How to use

You use the DeepL MCP Server by running the local server process and pointing your MCP client to it. The server exposes five tools for translation tasks: translate_text, rephrase_text, get_source_languages, get_target_languages, and get_usage. You can translate single texts or batches, control formality for specific languages, provide context to improve accuracy, and leverage glossary support when you need consistent terminology. Your MCP client can request the available source and target languages and monitor usage to stay within your plan limits.

How to install

Prerequisites you need before starting are Node.js 18 or newer (including Node.js 25) and a DeepL API key. Install and run steps are shown in practical sequence.

Additional steps and configurations

Use the following sample configurations to connect your MCP client to the local server. This demonstrates how to run the server locally and pass your API key via environment variables.

{
  "mcpServers": {
    "deepl": {
      "command": "node",
      "args": ["/path/to/deepl-mcp-server/build/index.js"],
      "env": {
        "DEEPL_API_KEY": "your-api-key-here"
      }
    }
  }
}

Claude Desktop configuration example

The following configuration illustrates how to load the local DeepL MCP Server into Claude Desktop. Adjust the path to point at where you build the server.

{
  "mcpServers": {
    "deepl": {
      "command": "node",
      "args": ["/path/to/deepl-mcp-server/build/index.js"],
      "env": {
        "DEEPL_API_KEY": "your-api-key-here"
      }
    }
  }
}

Claude Code configuration example

The following example shows how to configure Claude Code to load the local DeepL MCP Server from your project.

{
  "mcpServers": {
    "deepl": {
      "command": "node",
      "args": ["/path/to/deepl-mcp-server/build/index.js"],
      "env": {
        "DEEPL_API_KEY": "your-api-key-here"
      }
    }
  }
}

Usage overview

When you start the local server, you can call the five available tools through your MCP client. Translate text with target language and optional formality, request rephrasing (with Pro access where available), list source and target languages, and check the current API usage.

Development notes

The server is designed to auto-detect the API type based on your key format. Keys ending with ":fx" use the Free API endpoint, while other keys use the Pro API endpoint. Ensure your API key is kept secure and not exposed in client-side code.

Troubleshooting tips

If translations fail, verify that your DEEPL_API_KEY is correctly set in the environment, confirm network access to the DeepL API, and check that you started the server with the correct path to the built index. Review usage statistics to ensure you have remaining quota.

Notes

The server supports batch translation, preserves formatting, and allows you to control sentence splitting. You can also supply context for improved accuracy and use glossary IDs for consistent terminology.

Available tools

translate_text

Translate text with formality control, context, and glossary support

rephrase_text

Improve text clarity and style using DeepL Write (Pro only)

get_source_languages

List all available source languages

get_target_languages

List all target languages with formality support info

get_usage

Get API usage statistics (character count and limits)