Format × language

Translate a PDF document from Chinese to Korean, layout closely preserved

Need to translate a PDF document from Chinese to Korean without wrecking the formatting — or pasting sensitive data into a public translation box? Veil Translate parses the file in your browser, translates only the text, and writes it back into the original structure, so the result matches the source layout.

Sensitive entities are detected and masked locally before anything is sent, then restored on your machine after translation.

Translate my PDF documentFree · no sign-up · runs in your browser

What gets translated

Body paragraphs in the PDF are translated: the source text is removed from the file and the translation is placed at the original position, while page structure, images and figures stay put. Text-based PDFs only — scanned PDFs are not supported.

Sensitive data is masked locally

Before sending, the names, ID numbers, phones, emails, bank/USCC numbers and organizations you confirm are replaced with stable placeholders; only masked text goes to your chosen translation service, and restoration happens on your machine. The de-identified body text is still sent as plaintext — this is not encrypted translation.

Detection coverage & quality

Because the source is Chinese, sensitive entities are detected reasonably well by the local regex-plus-AI model and masked before translation. Placeholders are protected by the prompt and verified for integrity before restoration.

Frequently asked questions

Will the PDF document formatting change after translation?

Mostly no. Translations are placed back at the original paragraph positions, closely preserving the layout; long translations may use a slightly smaller font size.

Is my file uploaded to your server?

No. Parsing, detection, masking and restoration happen in your browser. Only masked text and your API key are sent, directly to the translation provider you choose.

Which sensitive values are detected?

Names, ID numbers, phones, emails, bank accounts, USCC numbers and organizations — by regex plus a locally-run AI model. You can add or remove any of them on the review screen before masking.

Related guides