Skip to content
DeepL

DeepL

v1.3.0
MIT
translationlocalizationwritingdocumentsglossary

Summary

DeepL's official MCP server: translate text and whole documents, rephrase copy with a chosen style and tone, and reuse your account's glossaries and style rules.

Features

  • Translate text with formality, context and formatting controls
  • Translate whole documents (PDF, DOCX, PPTX, XLSX, HTML, TXT) to disk
  • Rephrase text with selectable writing style and tone
  • Read account glossaries and their per-direction dictionary entries
  • Read configured style rules and custom instructions
  • Query the API for supported source and target language codes

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it, filling in the environment variables with your own values.
  3. Restart the agent — the "DeepL" tools will be available.
{
  "mcpServers": {
    "deepl-mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "deepl-mcp-server"
      ],
      "env": {
        "DEEPL_API_KEY": "your-api-key"
      }
    }
  }
}

Description

The official MCP server from DeepL exposes the DeepL API as thirteen tools, so an agent can translate, rephrase and look up your organisation's terminology without you leaving the editor.

Translation

translate-text handles inline strings with the full set of DeepL options — source and target language, formality, glossary, style rule, surrounding context, formatting preservation, sentence splitting and custom instructions. translate-document takes a real file (PDF, DOCX, PPTX, XLSX, HTML, TXT and more), translates it and writes the result to disk with the layout intact, which is the part that is awkward to do by pasting text into a chat window.

Rephrasing

rephrase-text rewrites text — optionally into another language — with a style and tone you pick from the values get-writing-styles and get-writing-tones report, so the agent can ask the API what is actually available instead of guessing.

Glossaries and style rules

This is what separates it from a generic translation tool: list-glossaries, get-glossary-info and get-glossary-dictionary-entries read the term bases already in your DeepL account, one language pair and direction at a time, while list-style-rules, get-style-rule and get-custom-instruction expose your configured writing rules and custom instructions. Translations stay consistent with the terminology your team has already agreed on.

Requirements

Node.js 18 or newer and a DeepL API key set as DEEPL_API_KEY; the server exits immediately without one. It communicates over stdio and is normally launched by the MCP client. Published as deepl-mcp-server under the MIT licence.

Related MCP Servers

MCP: Slack

by Zencoder

New

Read and post to Slack from an AI client — channels, threads, reactions, and user lookup.

Productivity
New

Official instance-level MCP server that lets AI agents discover, run, and edit workflows on your n8n instance.

Productivity
New

Home Assistant's built-in MCP Server integration — lets Claude, ChatGPT and other MCP clients see and control exactly the smart-home entities you choose to expose.

Productivity

MCP: Airtable

by Domenic Rosati

New

Read and write Airtable bases from an AI client — records, schemas, fields, and views.

Productivity
Browse all MCP servers →