How text extraction works
Inside a PDF, text is not stored as sentences and paragraphs. It is stored as drawing instructions: “write these letters, in this font, at this exact spot on the page”. A single line might be split into dozens of pieces, and the pieces are not always in reading order.
The tool reads those pieces with PDF.js, the PDF engine used by Firefox, and puts them back together:
- Pieces on the same baseline, the invisible line that letters sit on, become one line of text.
- A space is added when there is a visible gap between two pieces, but not when a word was merely split by a change of font or colour.
- When the baseline moves down by about one line, a new line starts. When it jumps by more than one and a half lines, a blank line marks a new paragraph.
The result is plain text with sensible line breaks, ready to paste into an email, a document or a spreadsheet. Everything runs in your browser, so the PDF is never uploaded.
Text PDFs vs. scanned PDFs
| Type of PDF | How it was made | Works here? |
|---|---|---|
| “Real” text PDF | Saved from Word, Google Docs, a website or an accounting system | Yes, quickly and accurately |
| Scanned PDF | A scanner or phone camera took a picture of paper | No: the words are pixels, not text |
| Mixed | A text document with a few scanned pages | Partly: the tool lists the pages with no text |
A quick test: open the PDF and try to highlight a sentence with your mouse. If you can select individual words, the text is real and this tool will extract it. If the whole page highlights as one block, or nothing at all, it is a scan.
Scanned pages need OCR (optical character recognition), which looks at the shapes of letters and guesses the words. Save the pages as images with PDF to JPG, then run them through Image to Text.
Worked example
You received a 3-page quote as a PDF and need the item list in a spreadsheet.
- Choose the PDF. The text appears within a second or two, with a line such as “742 words from 3 pages”.
- Untick Mark where each page starts, because you do not want “— Page 2 —” lines in the middle of the list.
- Select the lines you need in the box, or click Copy text to take everything.
- Paste into your spreadsheet. Each line of the quote lands in its own row.
If you want to keep the text for later, Download .txt saves it as quote.txt, which opens in any text editor.
Getting the cleanest result
- Columns: newsletters and academic papers often use two columns. The text usually comes out column by column, but occasionally lines from both columns are mixed; check and tidy as needed.
- Tables: cells on the same row are joined with spaces. For neat columns, paste into a spreadsheet and use its “split text to columns” feature.
- Hyphenated words: a word broken across two lines keeps its hyphen, just as it appears on the page.
- Headers and footers: page headers, footers and page numbers are part of the page text, so they appear too. Page markers make them easy to find and delete.
- Odd characters: a few PDFs use fonts without a proper character map, which produces gibberish. That is a problem in the file itself; OCR is the workaround.
What you can do next
- Check the length with the Word Counter.
- Compare two versions of a contract with Text Compare.
- Need pictures of the pages instead? Use PDF to JPG.
Privacy
The PDF is opened and read on your device. The extracted text exists only in this page until you copy or download it, and nothing is sent to a server.