Supports the file formats DeepL's document API accepts (.docx, .pptx, .xlsx, .pdf, .txt, .html). Formatting, layout and placeholders are preserved by DeepL automatically for these formats.
Works best on PDFs with a real text layer; scanned/image-only PDFs aren't OCR'd by DeepL
and may translate poorly or come back unchanged.
Only user-facing string values are sent to DeepL. Keys, numbers, booleans, nulls, URLs, HTML/code snippets,
template tags like {{var}}, and key-like identifiers (e.g. status_code_404) are
skipped and reconstructed verbatim — this is a heuristic filter, so review the output before shipping it.