DeepL document translation adds ten file formats in beta (Sep 29, 2026)

What you get

On September 29, 2026 DeepL added ten file formats to document translation, all in beta: xlsm (macro-enabled Excel), zip (SCORM e-learning packages), vtt (WebVTT subtitles), yaml / yml, properties (Java), strings (iOS and macOS), md / markdown, resx (.NET), odt and rtf.

Prerequisites

A DeepL API key in DEEPL_AUTH_KEY
The right address for your plan: https://api.deepl.com for Pro, https://api-free.deepl.com for Free

Steps

Upload the file to the same document endpoint as before. Only the file changes, for example a subtitle file:

Shell

curl -X POST https://api.deepl.com/v2/document \
  --header "Authorization: DeepL-Auth-Key $DEEPL_AUTH_KEY" \
  --form 'source_lang=EN' \
  --form 'target_lang=DE' \
  --form 'file=@subtitles.vtt'

Keep the document_id and document_key from the response. You need both to check on the translation and to download the result.
Check what is and isn't translated for your format:
- Subtitles (vtt): cue text is translated; timings, identifiers and the WEBVTT header are kept. Translated lines can be longer, so review how they fit.
- Markdown: paragraphs, headings, lists, blockquotes, table cells and image alt text are translated. Known issue: the --- lines around YAML front matter are dropped, so add them back.
- iOS .strings and Java .properties: values are translated, keys are kept. Placeholders are normally protected; check them afterwards.
- YAML: string values are translated at every level; keys, numbers, comments and indentation are kept.
- xlsm: cell text is translated; macros and VBA strings are left alone.

Test it

Upload a short .vtt file and download the result: the timings should be identical and only the spoken lines changed.

Not confirmed

The upload command is DeepL's own document example with the file swapped for a .vtt. DeepL's page does not show a sample for each new format.
These formats are in beta, so behaviour may change. For plain text in your app, see Translate your app's text with DeepL from Node or Python.

What you'll need

What you get

Prerequisites

Steps

Test it

Not confirmed