A PDF can look like an ordinary page while storing its words in an unexpected order, or as an image with no extractable words at all. Before translating, find out which kind of document you have. StarTranslate extracts text from a text-based PDF; it does not create a translated copy of the original page design.
Check the PDF before uploading
Open the document in a PDF reader and try selecting a complete sentence. Copy it into a plain-text editor. If the words appear in a sensible order, that passage is a useful starting point. Selecting a page-sized image, or obtaining an empty paste, suggests you may need optical character recognition first.
Selectable text is not a guarantee of clean extraction. A two-column brochure can mix sentences from separate columns. Headers, page numbers, captions, and footnotes may interrupt a paragraph. Compare the pasted passage with the visible page before trusting it.
Translate a short, coherent passage
The file must be no larger than 5 MB and a PDF must have 30 pages or fewer. Extracted text must fit within 5,000 characters. A small file can still exceed the character limit, so file size and text length are separate checks.
- Choose the file option and select your text-based PDF.
- Read the extracted text in the input box and correct broken words or reading order.
- Choose the source language explicitly if detection is uncertain, then select the target language.
- Pause for about one second after editing; translation starts automatically.
- Compare the result with the original passage, then copy or download the plain-text result.
Repair extraction problems before translating
For example, a brochure might extract “Opening hours Monday” from one column and insert an address from the next column before “to Friday.” Reconstruct the sentence from the original page rather than expecting translation to resolve the layout.
Remove repeated running headers and page numbers only when they are clearly separate from the content. Join a word split across a line carefully: a visible hyphen may be part of the word rather than a line-break artifact. Keep section headings when they explain what the following paragraph is about.
Know what the result contains
The download is a TXT file. Tables, columns, images, fonts, and original page positions are not preserved. If you need a formatted deliverable, place the reviewed translation into a separate document and check each label against its original position.
Scanned PDFs need an OCR step outside this tool. After OCR, inspect names and numbers especially carefully before pasting a short section. Do not bypass a document’s access restrictions. For official submissions, ask the receiving organization what translation format and certification it accepts.
Common questions
Can I translate a scanned PDF here?
Not directly. StarTranslate has no OCR feature. Obtain text through an appropriate OCR workflow, review it, then paste a section within the character limit.
Why does a short PDF exceed the limit?
The extracted text can contain more than 5,000 characters even when the file has only a few pages. Translate coherent sections instead of uploading the whole document.
Will the translated PDF keep its layout?
The result is plain text, not a newly generated PDF. Original formatting and document layout are not preserved.