Skip to main content

HTML translation

Set format: "html" to translate HTML while preserving markup.

How it works​

  • Langbly parses your HTML and translates text nodes only
  • Structural attributes (href, src, class, id, style, name, value, data-*) are preserved as-is
  • Content attributes that carry visible or accessible text (alt, title, placeholder, aria-label) are translated
  • script, style, and noscript contents are skipped

Example​

curl -sS 'https://api.langbly.com/language/translate/v2' \
  -H 'Content-Type: application/json' \
  -H 'X-API-Key: YOUR_API_KEY' \
  -d '{
    "q": "<p>Hello <strong>world</strong>.</p>",
    "target": "nl",
    "format": "html"
  }'

Frequently asked questions​

Why are alt and title attributes being translated after I read that attributes stay unchanged?​

Only structural attributes like href, src, class, id, style, name, value, and data-* remain untouched. Content attributes that users or screen readers see, alt, title, placeholder, and aria-label, are translated to the target language since they carry visible text.

Does the API translate placeholders like {{name}} or %s inside otherwise translatable text?​

No. Protected zones such as {{name}}, %s, %d, full URLs, email addresses, and @handles are never translated, even when they appear inside a text node that is otherwise processed.

What happens if my input HTML is malformed? Will the API reject the request?​

The parser handles invalid or incomplete HTML as best it can. There is no strict validation check that rejects a request for bad markup. The translation runs on whatever structure can be parsed.

Can I send HTML that contains <script>, <style>, or <noscript> blocks?​

Those tags and everything inside them are automatically skipped and never translated. Only visible text nodes and content attributes in other elements are processed.