Image to Text (OCR)
Extract editable text from screenshots, scanned documents, or photos — in English, Chinese, or both. The OCR engine runs entirely on your device, which is exactly why you can trust it with ID cards and contracts.
Drop an image here, click to choose, or paste from clipboard (PNG / JPG / WebP / BMP)
OCR is not perfect: accuracy depends on image quality, fonts and layout. Review the text before using it, especially for low-resolution, skewed or handwritten content.
Your image is recognized entirely in your browser — it never leaves your device. The OCR engine and language data are served from this site; no public CDN is contacted.
Something wrong with this tool? Tell us and we'll fix it
How to use
- Upload or paste an image containing text.
- Choose the language (English, Chinese, or auto).
- Copy or download the extracted text.
Frequently asked questions
How accurate is the OCR?
Very good on clean screenshots and scans (95%+ on standard fonts). Accuracy drops on handwriting, heavy compression, skew, or decorative fonts — crop tightly and use the clearest image you have.
Does it work for Chinese text?
Yes — select Chinese or bilingual mode. The engine handles simplified Chinese, English, and mixed documents.
Is my document uploaded?
No. Recognition runs locally in your browser with the Tesseract engine compiled to WebAssembly — the only download is the engine itself from a public CDN on first use.