DeepL MCP Server – Translation and Writing
DeepL's official MCP Server connects AI assistants to the DeepL API for translation and language workflows. Use it when an agent needs high-quality text translation, document translation with format preservation, rephrasing, supported-language lookup, or glossary-backed terminology control.
Overview
DeepL MCP Server is DeepL's official open-source Model Context Protocol server
for bringing DeepL translation capabilities into AI clients. It lets agents call
the DeepL API through standardized MCP tools, making multilingual workflows
available inside tools such as Claude Desktop, Claude Code, Cursor, GitHub
Copilot, and other MCP-compatible clients.
What the MCP server enables
DeepL MCP exposes tools for translation, writing improvement, language lookup,
and glossary workflows. Depending on the prompt and available files, an AI
agent can:
- Translate text between supported languages with automatic source-language
detection. - Translate documents such as PDF, DOCX, PPTX, XLSX, HTML, and TXT while
preserving document formatting. - Rephrase text using DeepL writing capabilities, including configurable style
and tone where supported. - List available source and target languages from the DeepL API.
- List glossaries available in the connected DeepL account.
- Retrieve glossary metadata and glossary dictionary entries.
- Apply a glossary ID to translation requests for consistent terminology.
When to use it
Use DeepL MCP when an AI agent needs reliable multilingual output or
localization help. Practical examples include translating a customer email into
German with formal tone, localizing a product brief into French, translating a
report document, rewriting an English paragraph in a professional style,
checking which target languages are available, or using a marketing glossary to
keep brand and product terminology consistent across translations.
Connection and authentication
The official quick-start command is npx deepl-mcp-server. Node.js v18 or
later is required. The server requires a DeepL API key supplied through the
DEEPL_API_KEY environment variable. DeepL documents Claude Code setup with
claude mcp add deepl -e DEEPL_API_KEY=your-api-key -- npx deepl-mcp-server
and Claude Desktop setup with a local MCP server configuration that runs npx
with the deepl-mcp-server argument.
A local installation is also supported with npm install deepl-mcp-server, and
development or source-based use can run the repository with Node.js after
installing dependencies.
Key considerations
DeepL MCP calls the DeepL API and therefore requires a valid DeepL API account
and key. Usage is subject to DeepL API billing, character limits, document
translation behavior, and account-level glossary access. Document translation
expects file paths, so the MCP client and agent need access to the relevant
filesystem tool or local files. Glossary-based translation requires a compatible
source language, target language, and glossary ID. Store DEEPL_API_KEY
securely, avoid committing it to configuration files, and verify translated
documents or legally sensitive text with a human reviewer before publication.
Supported Transports
stdio
Command: npx
Args:
deepl-mcp-server
stdio
Command: node
Args:
${DEEPL_MCP_SERVER_PATH}/src/index.mjs
Frequently Asked Questions
- When should an AI agent use the DeepL MCP Server?
- Use it when an agent needs to translate text, translate documents, improve writing, look up supported languages, or apply DeepL glossaries for consistent terminology in localization, customer communication, content, and documentation workflows.
- What does the DeepL MCP Server add to an AI agent's capabilities?
- It gives the agent access to DeepL API tools for text translation, document translation, writing rephrasing, source and target language discovery, and glossary lookup, instead of relying only on the model's built-in language abilities.
- What can an AI agent access or manage through DeepL MCP?
- The agent can translate text, translate supported document formats, rephrase text, list source and target languages, list glossaries, inspect glossary metadata, and retrieve glossary dictionary entries. It does not administer the entire DeepL account.
- How is authentication configured for DeepL MCP?
- The server requires a DeepL API key passed as the DEEPL_API_KEY environment variable. Store the key in the MCP client's secret or environment-variable configuration and do not commit real keys to source control or shared YAML files.
- Which transport should be used for DeepL MCP?
- Use local stdio with `npx deepl-mcp-server` for normal MCP client setup. Use the Node.js source path only when running a checked-out or locally installed copy of the official DeepL MCP server.