# DeepL MCP

> DeepL provides neural machine translation for over 30 languages. It goes beyond basic word swapping, capturing nuance and tone—whether your text needs to sound formal or casual. You can also build custom brand glossaries to guarantee consistent terminology across all translations, making it perfect for global marketing content.

## Overview
- **Category:** ai-frontier
- **Price:** Free
- **Tags:** machine-translation, language-processing, glossary-management, document-translation, api-integration

## Description

If you're building an application that requires multilingual support, the quality of translation matters more than anything else. This MCP connects your AI agent directly to DeepL’s translation engine. You stop worrying about whether a phrase sounds natural or if your brand terminology gets lost in translation. Instead, your agent handles it automatically.

This tool lets you translate standard text into dozens of languages while also controlling the formality—you can tell it to use a formal tone for corporate documentation and an informal tone for social media posts. The real power comes from managing custom glossaries; you define your brand terms once, and the system applies them everywhere, ensuring consistency across thousands of words. You'll manage everything, from creating those term lists to checking how much API quota you've used, all through Vinkius.

Ultimately, it lets developers integrate high-quality localization directly into their code or agent workflows without needing to build a massive translation layer themselves.

## Tools

### create_glossary
Sets up a new glossary containing specific word pairs for consistent terminology.

### delete_glossary
Removes an existing glossary when the branded term list changes or is no longer needed.

### get_document_status
Checks if a large document translation job has finished processing.

### get_glossary_entries
Lists all the specific terms and their translations stored within a glossary.

### get_glossary
Retrieves the details of an existing glossary, including its name and source language pair.

### get_usage
Provides detailed metrics on how many characters you've translated this billing cycle.

### list_glossaries
Shows a list of all the glossaries currently set up and available for use.

### list_glossary_language_pairs
Lists which language pairs are supported for using custom glossaries.

### list_source_languages
Provides a list of all languages that can be used as the source text.

### list_target_languages
Lists all possible destination languages for your translation output.

### translate_formal
Translates text specifically using a highly formal, professional tone.

### translate_informal
Translates text with an approachable, casual, and conversational tone.

### translate_text
Performs standard translation of text without specific controls for tone or glossary adherence.

### translate_with_glossary
Translates text while strictly following the custom rules and terminology defined in a loaded glossary.

## Prompt Examples

**Prompt:** 
```
Translate 'Welcome to our platform. We look forward to working with you.' into German (formal) and Brazilian Portuguese (informal).
```

**Response:** 
```
German (formal): 'Willkommen auf unserer Plattform. Wir freuen uns auf die Zusammenarbeit mit Ihnen.' Brazilian Portuguese (informal): 'Bem-vindo à nossa plataforma. Estamos ansiosos para trabalhar com você.' The formal version uses 'Ihnen' (Sie-form) while the informal Portuguese uses 'você' instead of 'o senhor'. Would you like to create a glossary for consistent terminology or translate more content?
```

**Prompt:** 
```
Create a glossary for EN→FR with our brand terms and then translate a marketing paragraph using it.
```

**Response:** 
```
Glossary created! ID: gl_8a4f2c, Name: 'Brand Terms EN-FR', Source: EN, Target: FR, Entries: 5 term pairs. Terms mapped: 'Dashboard' → 'Tableau de Bord', 'Workspace' → 'Espace de Travail', 'Insights' → 'Analyses', 'Onboarding' → 'Intégration', 'Workflow' → 'Flux de Travail'. Now translating with glossary... Result: 'Découvrez notre nouveau Tableau de Bord avec des Analyses en temps réel. L'Intégration est simplifiée grâce à notre Flux de Travail guidé dans votre Espace de Travail.' All 5 glossary terms were applied consistently.
```

**Prompt:** 
```
Check my DeepL API usage and list all available target languages.
```

**Response:** 
```
API Usage: 142,800 of 500,000 characters used (28.6%). Remaining: 357,200 characters. Plan: DeepL API Free. Billing period resets: May 1, 2025. Target languages available (31): BG (Bulgarian), CS (Czech), DA (Danish), DE (German), EL (Greek), EN-GB, EN-US, ES (Spanish), ET (Estonian), FI (Finnish), FR (French), HU (Hungarian), ID (Indonesian), IT (Italian), JA (Japanese), KO (Korean), LT (Lithuanian), LV (Latvian), NB (Norwegian), NL (Dutch), PL (Polish), PT-BR, PT-PT, RO (Romanian), RU (Russian), SK (Slovak), SL (Slovenian), SV (Swedish), TR (Turkish), UK (Ukrainian), ZH (Chinese). Languages with formality support: DE, FR, ES, IT, NL, PL, PT-BR, PT-PT, RU, JA.
```

## Capabilities

### Translate Text with Tone Control
Translates raw text while allowing you to specify if the tone should be formal, informal, or neutral.

### Enforce Brand Terminology
Applies custom glossaries to translations, guaranteeing specific company terms are translated the exact same way every time.

### Manage Language Pairs
Lists all supported languages and checks which language pairs can use a glossary for maximum flexibility.

### Monitor API Usage
Checks your current usage quota, showing how many characters you've consumed and when the billing period resets.

### Handle Documents
Tracks the progress of larger document translations submitted through the system.

## Use Cases

### Launching a new product line in Germany.
The marketing team needs to translate website copy into German. They use `create_glossary` first, inputting core terms like 'platform' and 'user dashboard'. Then, they call `translate_with_glossary`, ensuring the translations maintain the correct formal *Sie* address throughout the entire site.

### Updating social media copy for Brazil.
A content creator needs to take a blog post and adapt it for Instagram in Brazilian Portuguese. They bypass standard translation by calling `translate_informal`, guaranteeing the output sounds genuinely conversational, not robotic.

### Processing large technical manuals.
The documentation team submits an entire PDF manual. Instead of copy-pasting sections, they use the MCP to monitor the job status via `get_document_status`, and when complete, retrieve the full translated output.

## Benefits

- Stop worrying about tone. You can use `translate_formal` for official legal documents or `translate_informal` for social media posts, letting your agent pick the right register automatically.
- Never lose a brand term again. By setting up glossaries and using `translate_with_glossary`, you force consistency on key terminology, which is crucial for product names and jargon.
- It saves you time tracking API limits. The `get_usage` tool tells you exactly how many characters you've consumed versus your quota, so you never hit an unexpected billing wall.
- You can manage the entire process from one spot. Using tools like `list_glossaries` and `get_glossary_entries` means all your brand terms are organized and immediately ready for translation.
- Handles more than just text. If you submit a large file, the MCP monitors its progress via `get_document_status`, letting you know when the full document is ready.

## How It Works

The bottom line is that it gives your AI client access to high-quality, context-aware translation without complex setup or manual API calls.

1. First, you connect your DeepL API key to the MCP and initialize any necessary glossaries using tools like `create_glossary` or `list_glossaries`.
2. Next, your agent calls one of the translation methods—like `translate_with_glossary` for branded content, or `translate_formal` if you need professional copy—passing the text and target language details.
3. The system returns the translated text, which is accurate to the requested tone and adheres to any glossary rules you set up.

## Frequently Asked Questions

**Can I control the formality of translations (formal vs. informal)?**
Yes! Use `translate_formal` for professional communications (e.g., contracts, official correspondence) or `translate_informal` for casual content (e.g., social media, chat). The standard `translate_text` tool also accepts an optional formality parameter ('more', 'less', or 'default'). Note: formality control is available for select target languages including DE, FR, ES, PT-BR, and others.

**Can I create custom glossaries to ensure consistent terminology?**
Yes. Use `create_glossary` with a name, source language, target language, and TSV entries (tab-separated source→target pairs). Then use `translate_with_glossary` to apply the glossary during translation. Use `list_glossaries` to see all glossaries, `get_glossary_entries` to inspect term pairs, and `list_glossary_language_pairs` for supported combinations.

**How does DeepL authentication differ from standard Bearer tokens?**
DeepL uses a custom Authorization header format: `DeepL-Auth-Key YOUR_KEY` (not Bearer). Your API key is generated from the DeepL account dashboard. Free accounts use `api-free.deepl.com`, while Pro accounts use `api.deepl.com`. Use `get_usage` to check your current character consumption and plan limits.

**How can I check my API usage limits using the `get_usage` tool?**
You use `get_usage` to track your character count, remaining quota, and the billing period reset date. This feature keeps you informed about consumption so you don't hit rate limits unexpectedly.

**What if I send a large file? Can `get_document_status` help me monitor it?**
Yes, use `get_document_status` to track progress on submitted document translations. This tool provides updates until your job is complete and the translated files are ready for you.

**How do I confirm which languages are supported by listing them with `list_source_languages`?**
Calling `list_source_languages` or `list_target_languages` gives you a complete list of all supported language codes. The MCP provides quick access to 30+ options for your translation needs.

**How do I view existing custom terminology using the `get_glossary_entries` tool?**
You use `get_glossary_entries` and `list_glossaries` to view all saved terminologies. This lets you inspect exactly what was mapped before running a translation job.

**I just need basic text translated without formal or informal tone; should I use `translate_text`?**
Yep, `translate_text` is for standard translations that don't require specific tone settings. It handles basic text input across all supported language pairs smoothly and efficiently.