Free · offline · no sign-up
Image to text (OCR)
Drop in a screenshot or photo and get the text back out, in any of 18 languages including Chinese, Arabic, Japanese and Korean. The image is processed in your browser, which means it is never uploaded. That is a real difference if the screenshot has anything private in it.
Opens a fresh page with just this tool on it. Everything else is one click away.
How to use it
- 1 Drag an image in, paste from your clipboard, or pick a file.
- 2 Pick the language if it is not English, then wait a moment while the text is recognised.
- 3 Copy the extracted text and paste it wherever you need it.
Frequently asked questions
Is my image uploaded to a server?
No. Recognition runs locally in your browser, so the image never leaves your machine. That makes it usable for screenshots containing information you would not want to upload.
What kinds of images work best?
Clear, high-contrast images of printed or on-screen text work best. Screenshots are usually excellent. Handwriting and heavily stylised type are much less reliable.
Does it work for languages other than English?
Yes. Choose from 18 languages in the dropdown, including German, French, Spanish, Arabic, Chinese, Japanese, Korean and more, before you run recognition.
Can I paste a screenshot directly?
Yes. Copy a screenshot to your clipboard and paste it straight into the block.
It is not just this one tool
The image to text (ocr) is one block inside KontrolRoom, a modular workspace with notes, a calculating workbook, task boards and 20+ other utilities you can arrange on a single canvas. It all runs locally in your browser, works offline and costs nothing.
Explore the full workspace