Extract Text from PDF
Pull all selectable text out of a PDF into plain text you can copy or download, locally in your browser.
Loading Extract Text from PDF… If nothing happens, please enable JavaScript.
Extracting text from a PDF means reading the characters stored in the document and returning them as plain, editable text. It is one of the most frequently needed operations in any document workflow. You might want to quote a passage from a report, feed the contents of a paper into a search index, repurpose copy from an old brochure, count words, or simply get clean text out of a file that does not allow easy selection in your usual viewer. Rather than retyping anything, you let the tool walk through the document and hand you everything it can read.
Frequently asked questions
Are my files uploaded to a server?
Why does a scanned PDF return little or no text?
How is the extracted text organized?
Can I copy the text or save it to a file?
Does it preserve formatting like bold, tables, or columns?
What is the maximum file size or page count?
Does this work with password-protected PDFs?
Does it handle non-English text and special characters?
Can I extract text from many PDFs at once?
About Extract Text from PDF
This tool uses pdf.js, the same engine that powers PDF viewing in modern browsers, running entirely on your device. For each page it requests the text content and joins the individual text fragments together with spaces, then separates pages with blank lines so the output stays readable and roughly mirrors the layout of the original. The result appears in a read only text box along with the page count, and you can copy the whole thing to your clipboard with one click or download it as a .txt file ready to open in any editor.
Everything happens in your browser, with nothing uploaded. The PDF is read from your local disk and processed in memory, which keeps private contracts, research, and internal documents off any third party server. One important limitation to understand is that this tool reads text that is actually stored as text in the file. A scanned document or a photo saved as a PDF contains only images of words, with no underlying character data, so it will return little or no text. For those files you would need optical character recognition, which recognizes letters from the pixels, a different process from the direct extraction performed here.
Why Getting Text Out of a PDF Is Harder Than It Looks
A PDF does not store text the way a word processor document does. Instead of sentences and paragraphs, it stores drawing instructions that say, in effect, place this glyph at this exact coordinate on the page. There is often no explicit space character between words and no notion of where one line, column, or paragraph ends and the next begins. Extraction software has to reconstruct readable text by looking at the positions of the glyphs, inferring spaces from the gaps between them, and guessing reading order from their coordinates. This is why text pulled from a complex layout can sometimes arrive with words in an unexpected order or with spacing that does not match what you saw on screen.
The situation is complicated further by fonts. Each font in a PDF maps the codes in the content stream to glyph shapes, but the link back from a glyph to its actual Unicode character is optional, carried in a structure called a ToUnicode map. When that map is present, extraction is clean. When a PDF uses a subset or custom font and omits or mangles the map, the extracted output can be gibberish even though the page looks perfect, because the viewer knows how to draw the shapes but the file never recorded which characters they represent.
Then there is the great divide between real text and pictures of text. Around the world, an enormous share of PDFs are scans: photographs or flatbed images of paper, wrapped in a PDF container. To a human they look identical to a born-digital document, but they contain no character data at all, only pixels. Reading them requires optical character recognition, a field with roots stretching back to the 1920s and 1930s and devices built to help blind readers and to sort mail. Modern OCR uses machine learning to achieve high accuracy across many languages, but it remains a fundamentally different and more error-prone task than simply reading the text a digital PDF already contains, which is exactly what this tool does.