Sometimes you want the words and nothing else — no columns, no fonts, no layout to fight with. Upload your PDF above and get a plain .txt file containing the text, ready to search, paste, script against or archive.
One check decides whether this will work at all, and it takes ten seconds. Below: what that check is, why line breaks land where they do, and the encoding setting that determines whether your accents survive.
Open the document and try to select a sentence with your cursor.
If the text highlights, the PDF contains real text and extraction will work directly.
If nothing highlights, or the whole page selects as one block, your PDF is a scan — a photograph of a page. There is no text inside it, only pixels arranged so your eye reads them as words. Extraction will return an empty file, and no converter can change that.
Scans need OCR PDF first, which recognises the letter shapes in the image and produces text from them. Two things worth knowing: OCR makes mistakes, so the output always needs a read-through; and it works far better on straight, high-contrast pages, so run Rotate PDF first if the scan is sideways.
This is the most common complaint about extracted text, and it follows directly from how PDFs work.
A PDF does not store paragraphs. It stores instructions to place characters at particular coordinates on a page. The paragraphs you see are an illusion created by where those characters happen to sit — there is no marker in the file saying "this sentence continues from the line above."
So extraction has to infer paragraph structure from position, and it errs on the side of preserving what it can see. The usual result is a line break at the end of every visual line, mid-sentence, exactly where the original page wrapped.
Two ways to deal with it:
Worth expecting rather than discovering.
Multi-column layouts interleave. An academic paper or a newsletter with two columns can produce text that alternates between them, because nothing in the file states which column to read first — the characters simply sit side by side on the page. Extracting each column separately gives a far better result than untangling the output afterwards.
Tables lose their grid. Cells become a stream of values separated by spaces or tabs, and the row-and-column relationship disappears. If the data matters more than the prose, PDF to Excel attempts to rebuild the structure rather than flattening it.
Headers, footers and page numbers arrive mixed into the body. They are page content like anything else, so a running header appears every few hundred words throughout your file. Easy to strip with Find and Replace once you spot the pattern.
A .txt file carries no record of which character encoding it uses, so software opening one has to guess. Guess wrong and accented letters turn into strings like é, and non-Latin scripts become unreadable.
Extracted text is UTF-8, which covers every language. Most modern editors detect this correctly. Older versions of Windows Notepad sometimes do not — if your French, Arabic or Chinese text looks corrupted, open it in a different editor and explicitly select UTF-8 before concluding the extraction failed.
One related quirk: PDFs often use typographic characters that look like ordinary ones but are not — curly quotation marks, en and em dashes, non-breaking spaces, ligatures where "fi" is stored as a single character. They extract faithfully and then break things downstream if your destination expects plain ASCII. Find and replace them if you are pasting into code or a strict import format.
Stripping a document back is not only a loss. Several jobs get much easier once the words are separated from the page.
Searching across many files. Once documents are text, your operating system's search or a simple command-line tool can scan hundreds of them at once for a name, a figure or a phrase.
Comparing two versions. Extract both to text and run them through any diff tool, and every wording change appears immediately — far more reliable than reading two PDFs side by side.
Word counts for translation or editing. Translators and editors quote per word, and a PDF gives you no straightforward way to count.
Feeding another system. Scripts, search indexes, text-analysis tools and translation software want raw characters, not a page-layout container.
Long-term archiving. Plain text is the most durable format there is — a .txt file written decades ago still opens today, which cannot be said of most document formats.
Working with a scan? OCR PDF makes it readable first. PDF to Word keeps the formatting, PDF to Excel suits tables, and TXT to Word turns the result back into a document. Browse all PDF tools.