PDF 轉文字

The PDF is most likely a scan or a photo, which has no text layer to extract. Run it through OCR PDF first; that tool adds a text layer and also offers a .txt download directly.

Text is read left to right, top to bottom, following the PDF's internal structure, which is correct for most single-column documents. Multi-column layouts, sidebars, and tables can come out interleaved, and headers, footers, and page numbers are included with everything else on the page.

No. The output is plain UTF-8 text, so accented characters, Cyrillic, CJK, and Arabic are preserved, but bold, headings, links, and images are not, and line breaks follow the PDF's lines rather than paragraphs. For headings and lists kept as Markdown, use PDF to Markdown.

Yes. Add as many PDFs as you like and click Extract once; each becomes its own .txt file named after the source, and they download together as pdf-to-text.zip. A single PDF downloads directly as filename.txt.

Pages follow one another with a line break between them. There are no page markers or form feeds, so the file reads as one continuous stream of text.

No, this tool always extracts every page. Pull the pages you need into a separate PDF first with Extract Pages, then convert that file.

If a PDF asks for a password when you open it, the tool prompts you for that password before extracting. Without the password, the text can't be read.

No. Extraction runs in a PDF engine loaded into your browser as WebAssembly, so contracts, medical records, and anything else stay on your device from start to finish.