How to use it
- Choose a PDF of up to 40 pages.
- If a page is a scan, pick the language that matches it.
- Download the Word file. A longer line can overlap. A shorter line leaves a gap.
A PDF made by a program already stores each line’s place and size. This tool writes those lines into Word on the same page size, with bold or italic when the font name says so. A photo of the page is not included, because that picture was covering the words.
Word may substitute a font, because a PDF often embeds only part of one, so a line can end a little earlier or later. A table stays as placed words, not as a Word table that reflows. Changing a sentence does not push the next line down.
A scanned page has no words until recognition runs. Pick a language. The first run can download language data. That download is the recognition model, not your PDF. Proofread names and numbers. Handwriting and tiny print can be wrong. A scan stops at 20 pages so a phone can finish.
Questions
Will the Word page match the PDF?
The page size and the positions of the lines are kept, with the font size and bold or italic when the PDF font name includes them. Word may substitute a font, so a line can end slightly earlier or later.
Can I convert a scan?
Yes. A page with no text is recognized in the language you pick, up to 20 pages. The Word file gets those words, not a picture of the page. Proofread the result.
Does the file upload?
No. Recognition language data may download once. The PDF stays in the browser.