Add an image
Choose a screenshot, photo or scanned image. The browser prepares a local working copy for OCR.
OCR / Image to Text
Turn screenshots, photos and scanned images into editable text directly in your browser with OCR support for 30 languages.
Private
Your image stays on your device.
First run
Loads the OCR runtime and selected language data.
Later runs
Cached files and a reused worker make repeated OCR faster.
Drop an image or screenshot here
JPEG, JFIF, PNG, WebP, HEIC, HEIF, AVIF, SVG, BMP and TIFF · max 20 MB · max 7000 px source edge · about 30 MP.
OCR, or optical character recognition, converts visible text inside an image into text that you can copy, edit or save. It is useful for screenshots, photographed notes, scanned documents, labels, printed pages and other situations where the words are trapped inside an image instead of being selectable.
Image Toolroom performs OCR in the browser. The OCR engine and selected language data are loaded as local processing resources, while the image itself stays on your device for recognition. You can choose the language combination that best matches the source and adjust recognition options for different page layouts.
Choose a screenshot, photo or scanned image. The browser prepares a local working copy for OCR.
Select the language or languages that match the text and adjust page segmentation or rotation options when needed.
Run recognition, correct any OCR mistakes, copy the text or download the extracted result as a text file.
OCR stands for optical character recognition. It analyzes the shapes of printed or displayed characters in an image and converts them into machine-readable text.
This tool is primarily intended for printed or clearly rendered text. Neat handwriting may sometimes produce partial results, but handwriting recognition is much less predictable than printed-text OCR.
Recognition quality depends on resolution, focus, contrast, font style, page angle, language selection and image noise. Similar-looking characters such as O and 0 or I and 1 are common sources of mistakes.
Yes. When the image contains multiple languages, you can select the relevant OCR language combination. Loading more language data can make initialization slower.
The normal OCR workflow runs locally in the browser. The image and recognized text do not need to be sent to an Image Toolroom processing server for recognition.