OCR — scanned PDF to text
Recognise the text in a scanned PDF and get it back as .txt or .docx.
About this tool
A contract that was signed then scanned, an old digitised file, a receipt photographed on a phone: these PDFs carry no text layer, so nothing can be searched or copied. Optical character recognition reads the pixels and gives you words back. Drop the document: every page is rendered at high resolution then analysed, with real page-by-page progress, and the recognised text is shown page by page.
The engine runs in your browser, not on a server: a contract or a payslip never leaves your device, which matters a great deal for confidential paperwork. The document language is preselected from the site language and can be changed in one click — seven languages available. The result downloads as plain .txt or as a .docx you can edit straight away in Word. Average confidence is shown honestly: a pale or skewed scan scores low, and it is far better to know that now than to discover the mistakes later.
Frequently asked questions
How do I know whether my PDF needs OCR?
The tool checks as soon as you drop the file. If it already finds selectable text, it tells you and points you to PDF to Word instead: direct extraction is instant and perfectly faithful, where OCR has to guess characters. If the pages are pure images, recognition starts on its own.
Are my documents uploaded anywhere?
No. The recognition engine and the language data are downloaded once into your browser, and every calculation then happens on your machine. No page, no image and no recognised text is ever sent to a server.
Why does recognition take a while?
Expect a few seconds per page: your own device analyses every character, with no server farm behind it. The first run also downloads the language data, a few megabytes, cached afterwards. You can stop midway and keep the pages already processed.
The output has mistakes — what can I do?
First check that the right language is selected: French text analysed as English comes out as gibberish. After that, scan quality decides everything — a document scanned at 300 dpi, flat and well contrasted, beats a photo taken at an angle almost every time.
Is it really free, with no sign-up?
Yes — the tool is free, needs no account and has no usage limit. Processing happens locally in your browser, so your data never leaves your device.