Read English text from a screenshot locally, edit the result, then copy it or download a .txt file. First use downloads ~7 MB (engine ~3.9 MB + English data 2.9 MB + JavaScript). Inputs are limited to 25 MP. Your image stays on your device.
Choose an image to get started.
FAQ
How do I get the text?
Choose a clear English screenshot and follow recognition progress. Review and edit the text area, then choose Copy text or Download .txt. Small lettering and complex layouts may need corrections.
Does my screenshot leave my device?
No. The self-hosted recognition library, worker, core and English data load only when this tool needs them. The worker terminates after recognition. Viewing counts nothing; each successful copy or .txt download counts once, failures count zero.
Which licences cover OCR?
Tesseract.js and Tesseract.js-core use Apache-2.0, and the English tessdata also uses Apache-2.0. Licence copies are shipped under /vendor/tesseract/ and /vendor/tesseract/lang/LICENSE-tessdata.