Doc2Lang

Layout-preserving translation is the reason people pick Doc2Lang over a chat window: the file comes back intact. Translate a Word document here and headings, tables and headers and footers stay put. Translate an Excel file and merged cells, currency formats and charts survive the trip. Translate a PDF and the multi-column layout holds, and you get three versions at once: background preserved, text only, and a bilingual file with original and translation on alternating pages. PowerPoint keeps slide designs, while speaker notes travel along with the slide text.

Nothing is charged until you've seen something. After upload, Doc2Lang translates a portion of the document and shows it beside the source, so you can judge tone and terminology before committing. The exact price for Lite and Pro appears at the same time, calculated from the file itself rather than from its size in megabytes. If the preview is wrong, you can switch models and regenerate it before paying.

Lite and Pro are the same engine at different settings and prices. Lite starts at $2.90 per file and runs faster, which suits routine documents where a small wording slip won't hurt. Pro starts at $5 and is meant for contracts, specialist terminology and marketing copy where small word choices carry meaning. The company says text documents can cost up to 80 percent less on Lite than on Pro. Audio, video and scans narrow that gap because transcription and OCR add work.

Scanned PDFs and image-only files are read with optical character recognition before translation. That covers photographed contracts, printed manuals and pages where the text layer is unusable. Recognition quality depends on the source, so a clean 300 dpi scan behaves very differently from a phone photo taken at an angle. Low-quality scans should be checked page by page.

Beyond documents, Doc2Lang handles MP4, MOV and MKV video by extracting the audio, transcribing speech and translating the subtitles before re-timing them to the new text. MP3, WAV and M4A files go through the same transcription path. For images, the platform detects text inside JPG, PNG and WebP files, translates it and redraws it back onto the artwork, and you also get a text copy of the result.

You can drop a single ZIP archive of up to 200MB and pick files from it one at a time. The archive is unpacked in your browser, not on the server, so only the files you explicitly upload leave your device. Settings for language, model, style and glossary apply to the whole archive, and each file is billed separately when you send it.

Text documents accept a document description, a style preset and a glossary. That combination matters for anything with repeated proper nouns. Pin the company's official product names in a glossary and they translate the same way on every page, while the style setting shifts the register toward general, formal, casual, academic or technical.