Image to text (OCR)
Drop a photo, a screenshot or a scanned PDF and get the text, selectable and searchable. Recognition runs on your device; nothing is uploaded.
Updated November 6, 2026
How it works
- Drop an image or a PDF (scans are fine).
- Pick the language of the text; add English for mixed documents.
- Extract the text: copy it, save it as .txt, or build a searchable PDF.
What it reads well and what it does not
Printed text is its job: a scanned letter, a screenshot of an article, a photo of a page, an invoice, a receipt, a book page. A sharp, straight, evenly lit picture reads almost perfectly; a photo taken at an angle in a dim room loses words, and a crop around the text with Crop image before recognition helps. Handwriting is not read, nor are stamps or signatures, and a decorative font may come out with odd letters. Pick the language of the text from the list and add a second one for a document that mixes two.
Layouts and output
Auto finds columns and tables on its own. One block of text suits a single paragraph or a page of prose; One line is for a serial number or a code; Scattered text for labels and numbers placed anywhere. Keep column spacing keeps the alignment of a receipt or a table, and Lines as read keeps the layout where Paragraphs joins the lines and undoes hyphenation. Drag a rectangle over the picture to read just that part, for one address on a full page.
Scanned PDFs and what is downloaded
A scanned PDF of up to 200 MB is read page by page; type a page range to do only part of it, and pages that already carry text are taken as they are. Also build a searchable PDF gives the same pages with an invisible text layer on top, so the scan still looks like the original but can be searched and copied. The recognition engine, about 5 to 7 MB, is fetched once and kept, and the work runs in your browser; the document never leaves your device.
Why people use it
Free, no account
Every tool is free, with no sign-up, no e-mail address and no limit on how often you use it.
Your files stay on your device
The browser opens the file and processes it in memory, on your device. Nothing is uploaded, so nothing can be kept, leaked or looked at.
No watermark
The result is yours as it is: no watermark, no branding, no trial version.
Any device
A phone, a tablet or a computer with a current browser is enough: no app to install, no software to update.
Works offline after the first visit
The browser keeps the pages and their engines, so a tool you have opened once keeps working without a connection.
Open-source engines
The code that does the work is open-source software, named on the Licences page, running as WebAssembly in your browser.
Why it is private
Everything you type or drop here is handled by your browser, on your device. Nothing is uploaded. To verify, open your browser's developer tools, go to the Network tab, and use the tool: you will see no request carrying your data.
While a file is being processed, the counter under the result shows the page's network traffic: 0 requests and 0 B for your data.
How to check it yourself, step by step
What you can check yourself
Open your browser's developer tools on the Network tab before you drop a file: while the tool works you will see no request carrying your data, and the counter under every result shows the page's own measurement.
The site sets no cookies, so there is no cookie banner. Anonymous usage counts and error reports without file names are the only things that leave the page, as the privacy policy describes.
Every change goes through thousands of automated tests in three browser engines, including a guard that fails a test if a single byte of a file leaves the page.
Frequently asked questions
- How accurate is it?
- Tesseract reads clean, straight scans and screenshots very well. Low light, skew, handwriting and decorative fonts lower the accuracy; a sharper photo helps more than anything.
- Which languages are supported?
- English, Romanian, Spanish and Portuguese, alone or combined with English. Each language model is 1.5 to 3 MB and is fetched from this site the first time you use it.
- What is a searchable PDF?
- The same pages with an invisible text layer on top: you can search, select and copy the text while the scan still looks like the original.
- Does the file leave my device?
- No. The OCR engine (Tesseract compiled to WebAssembly) is loaded from this site and runs in your browser; the image and the text stay on your device.
- Does it read handwriting?
- No. It reads printed and typed text only; handwritten notes, signatures and stamps come out as nothing or as random letters.
- Can I use a photo of a document taken with my phone?
- Yes, and the tool runs in the phone's browser too. Hold the phone flat above the page in good light, keep the text straight, and crop away the table around it for the best result.
- Can I copy the text out of a screenshot?
- Yes. Paste or drop the screenshot, choose the language and press Extract text; the text appears ready to copy or to save as a .txt file.
- Does it handle tables and receipts?
- Partly. Choose Lines as read and Keep column spacing to keep the rows and columns aligned as plain text. The result is text, not a spreadsheet, so columns may need tidying.
- Can I read many images at once?
- One file at a time. For a batch, put the images into one PDF with JPG to PDF first and drop that PDF here; every page is read in one run.
Install it as an app
Chrome, Edge and Android browsers offer to install this site from the address bar or the menu; on iPhone use Share, then Add to Home Screen. It opens like an app, keeps working offline and still processes everything on your device.