DeepL

Getting Started

DeepL is a neural machine translation service known for producing highly accurate and natural-sounding translations in over 30 languages. This guide is for anyone who wants to use DeepL—whether through the web interface, desktop app, or API—including developers, content creators, and business professionals seeking reliable, context-aware translations.

Setup & Registration

Web App (Free & Pro)

1. Go to deepl.com and click Sign up free.
2. Provide your email, create a password, or use Google/Apple SSO.
3. Verify your email (check spam folder) – no further wait time.
4. After login, you can start translating immediately. For Pro features (e.g., glossary, data privacy), upgrade via account settings.

API Key (for Developers)

1. Log in to your DeepL account, then navigate to Account → API Keys.
2. Click Create a new API Key. Choose Free or Pro plan.
3. Copy the generated key (e.g., d4e5f6...). Save it securely – you won’t see it again.
4. The free tier allows 500,000 characters/month and no concurrent requests. Pro starts at $8.99/month with higher limits.

Your First API Call / First Use

Using the Web Interface

Simply paste text into the left box, select source/target language, and see the translation instantly. No code needed.

Using the API (cURL)

Replace YOUR_AUTH_KEY with your actual API key. This example translates "Hello, world!" from English to German.

curl -X POST 'https://api-free.deepl.com/v2/translate' \
  --header 'Authorization: DeepL-Auth-Key YOUR_AUTH_KEY' \
  --data-urlencode 'text=Hello, world!' \
  --data-urlencode 'target_lang=DE'

Expected response:

{
  "translations": [
    {
      "detected_source_language": "EN",
      "text": "Hallo, Welt!"
    }
  ]
}

Using the API (Python)

Install the official deepl library:

pip install deepl

Then run:

import deepl

translator = deepl.Translator("YOUR_AUTH_KEY")
result = translator.translate_text("Hello, world!", target_lang="DE")
print(result.text)  # Output: Hallo, Welt!

Common Use Cases

1. Localizing a Website or App

Use the API to translate dynamic content on the fly. For example, translate a product description for a German user:

product_desc = "This lightweight backpack is perfect for hiking."
translated = translator.translate_text(product_desc, target_lang="DE")
# Output: Dieser leichte Rucksack ist perfekt zum Wandern.

2. Translating Customer Support Responses

Integrate DeepL into a support ticket system. Example JSON response for a Spanish user:

curl -X POST 'https://api-free.deepl.com/v2/translate' \
  --header 'Authorization: DeepL-Auth-Key YOUR_KEY' \
  --data-urlencode 'text=Your order has been shipped.' \
  --data-urlencode 'target_lang=ES'

3. Document Translation (Web App)

Upload a .docx, .pptx, or .pdf file directly on the DeepL website. The translated document preserves formatting and layout. This is ideal for legal contracts or marketing materials.

Tips & Best Practices

Stay within Free Tier Limits

• Monitor your character usage via the DeepL API usage dashboard.
• For large volumes, batch translations and cache results to avoid repeated API calls.
• Use the split_sentences parameter to avoid exceeding character limits on single requests.

Common Pitfalls to Avoid

• Do not hardcode API keys in public repositories. Use environment variables.
• Do not translate extremely long texts in one API call – split into paragraphs (e.g., 10,000 characters per request).
• Do not forget to specify source_lang when the language is ambiguous; otherwise DeepL will auto-detect, which may be wrong for short texts.
• Free API does not support concurrent requests – sending more than one request at a time will result in errors.

Performance Optimization

• Use the formality parameter to control tone (e.g., prefer_more for formal business translations).
• For repeated translations (e.g., same phrases), create a glossary via the API or web interface to enforce specific terminology.
• Use the desktop app (Windows/Mac) for offline translations (Pro plan) or for large files without internet latency.
• For high-volume API usage, upgrade to the Pro plan to get faster response times and higher character limits.