Translate PDF English to Japanese

Yes. DocTranslator embeds CJK-capable fonts in the output, so every Japanese character renders properly instead of showing blank boxes (tofu). This matters most for scanned PDFs and older files whose original Japanese font is no longer available.

The layout is preserved, including vertically set text (tategaki), multi-column pages, tables, and the position of seals and figures. The translated page mirrors the structure of the original instead of being dumped into a single text column.

The AI reads the surrounding context to choose a fitting register, so business and formal documents come out in suitably polite Japanese rather than a flat, literal rendering. For high-stakes legal wording, have a native reviewer check the final text.

The pair works in both directions on the same engine, but the output behaves differently. English set from Japanese usually runs longer than the source, because Japanese packs a great deal into a few characters, so lines that fitted a narrow column may need more room. Two details are worth checking by eye afterwards: personal names, which appear surname first in Japanese and are often reversed in English, and dates written in era years, where Reiwa 6 and Heisei 31 mean nothing to a reader expecting 2024 and 2019.

Not on its own. A koseki extract is the usual stumbling block: the family registry is a structured record of a household rather than a single certificate, and the receiving office needs a human translator to sign a statement that the rendering is accurate and complete. Immigration filings and courts ask for that certification regardless of how good the translation reads. Our certified translation option covers that route.

Because of the fonts. A Latin font carries a few hundred glyphs while a Japanese one has to carry thousands, so embedding it inflates the file well beyond what the page count suggests. A twenty page Japanese document can weigh more than a hundred page English one. The 1 GB limit on the Pro and Enterprise plans is generous enough that this rarely matters, but it does explain why file size is a poor guess at how much text you are sending.

AI translation is $0.005 per word on the Free and Storage plans, and a page of Japanese gives you no reliable way to count those by looking at it, since the writing runs continuously without spaces. The practical answer is to measure rather than estimate: translate one short document first, which costs from $0.99 and shows what a real document of yours actually comes to. Pro lowers the rate to $0.004 per word for $49.99 a month, but no plan includes translations, so each document is still paid for. See pricing for the full detail.

Image files (JPG, JPEG, PNG) and image PDFs are supported, and a clean high-resolution scan of printed text reads well. Three things make Japanese scans harder than most. Furigana, the small reading aids printed above or beside kanji, sit close enough to the main line to be picked up as part of it. Vertically set pages have to be read in the right direction before anything else can happen. And handwriting is a different problem altogether, because cursive kanji bears little resemblance to its printed form. Where you have the original digital file, it will always beat a photograph of the page.