PDF to Text

How do I convert a PDF to text?

Press Choose PDF files and pick one or more PDFs. The text appears in the box within a second or two. Choose whether to join lines into paragraphs and remove repeated headers, then press Copy text or Download .txt. For several PDFs, the download is one ZIP with a text file for each.

1. Press Choose PDF files.
2. Check the text in the box; switch files with Showing if you picked several.
3. Set Join lines into paragraphs and Remove repeated headers, footers and page numbers.
4. Press Copy text, or Download .txt.

How do I copy text from a PDF without the line breaks?

When you copy from a PDF reader, every line ends with a hard break and words split by hyphens stay split. This tool rebuilds paragraphs from the positions of the lines: it joins lines that belong together, removes line-end hyphens inside words, and starts a new paragraph at wider gaps and indents. Untick Join lines to keep the original breaks.

Why does my PDF give no text, or strange symbols?

No text usually means the PDF is a scan: each page is a photo, with no letters stored. That needs OCR (optical character recognition), which this tool does not do yet. Strange symbols mean the PDF uses a font with no map back to letters, so readers cannot tell what each shape is. The tool warns you in both cases.

What is the JSON option for?

JSON gives the text in a structured form: one entry per page with its page number, and each block marked as a heading, paragraph, list item or code. It also includes the PDF's title, author and dates when the file has them. That suits scripts, databases and AI tools that need to cite a page.

{
  "source": "report.pdf",
  "pageCount": 12,
  "pages": [
    { "page": 1, "text": "…",
      "blocks": [{ "type": "heading", "level": 1, "text": "Annual report" }, …] }
  ]
}

How accurate is the extracted text?

For PDFs made by a word processor or browser, the words come out exactly as stored. Order can differ in complex layouts, such as text boxes or tables. We ran the tool on test files with known content and counted results.