Extract Text from PDF

Copying text straight out of a PDF viewer usually goes wrong: every line ends with a hard break, words split by a hyphen stay split, and two-column pages come out as left-right-left-right nonsense. This tool reads the text the PDF actually contains and rebuilds it properly before you copy it.

In Flowing paragraphs mode the lines of each paragraph are joined back together, hyphenated words are mended (“infor-” + “mation” becomes “information”), side-by-side columns are read one after the other, lists keep their bullets and numbers, and table rows are separated by tabs so they paste neatly into a spreadsheet. Keep layout mode does the opposite: every line stays where it sits on the page, with spaces for the gaps, which suits receipts, forms and wide tables.

You see the result page by page, can copy a single page or all of it, pick a page range, and save it as a plain .txt file or as Markdown (.md) with headings, bold and italic kept. If a page has no text at all — a scan or a photo — the tool says so instead of handing you a blank file, and points you to PDF OCR to read it. Everything happens in your browser, so the PDF never leaves your device.

1. Tap “Choose a PDF” or drop one onto the box. The text appears below, one card per page.
2. Pick “Flowing paragraphs” for reading and pasting, or “Keep layout” to keep columns and spacing as on the page.
3. Optional: type the pages you want in the Pages box, such as 2-5, 8, and use Show to look at one page at a time.
4. Press “Copy page” on any card, or “Copy all text” to take everything.
5. Press “Download .txt”, or “Download .md” for Markdown with headings and bold kept.

• Two modes: Flowing paragraphs (lines joined, hyphens mended) and Keep layout (spacing kept).
• Reading order for two- and three-column pages, with full-width titles kept in place.
• Bullet and numbered lists kept; simple tables come out tab-separated.
• Per-page cards with a word count and a Copy page button, plus Copy all text.
• Page range box (1-3, 7, 10-) and a Show picker for one page at a time.
• Downloads as .txt, or .md with headings by font size, bold and italic.
• Optional “--- Page 3 ---” headings in the download, so you can tell pages apart.
• Spots pages with no text (scans and photos) and links to PDF OCR.
• No upload, no sign-up, no page limit and no watermark.

• For a table you want in a spreadsheet, PDF to Excel gives you real rows and columns rather than tabs.
• If every page says “no text”, the PDF is made of pictures. Run it through PDF OCR first, then open the new file here.
• Strange symbols instead of letters mean the PDF was made with a font that hides its character map; copying from any app gives the same result, and OCR is the fix.
• Keep layout uses a monospaced view so columns line up; paste it into a plain-text editor, not a word processor, to keep the alignment.
• Headings in the Markdown file are guessed from text size, so a document that uses only one size will have none.