Does this OCR upload my image?
No. The selected image and extracted text are processed in this browser. The page downloads the public OCR engine and selected language model when you start, and the browser may cache those public files.
Tools
Extract editable English, Traditional Chinese, or Simplified Chinese text from a JPG, PNG, WebP, or screenshot in your browser.
The image and extracted text stay in this browser. The OCR engine and selected public language model load only after you start.
Drop one JPG, PNG, or WebP image here
You can also paste an image from the clipboard while this tool is open.
One image, up to 10 MB, 12 megapixels, and 8,192 px on either side.
Still JPG/JPEG, PNG, and WebP only. Animated images and files whose extension or MIME type conflicts with their contents are rejected.
Keep this tab open and the screen awake. A mobile browser may pause recognition when it is backgrounded or locked.
This image-to-text tool runs OCR in your browser after you choose a language and start recognition. It is useful for printed notices, clear screenshots, scanned pages, labels, and other horizontal text. The result is plain, editable text, so you can correct OCR mistakes before copying it or downloading a UTF-8 TXT file.
No. The selected image and extracted text are processed in this browser. The page downloads the public OCR engine and selected language model when you start, and the browser may cache those public files.
The tool accepts one still JPG/JPEG, PNG, or WebP image. It rejects GIF, animated PNG/WebP, HEIC, AVIF, TIFF, BMP, SVG, and PDF.
Yes. Choose Traditional Chinese + English. That mode is intended for horizontal printed Traditional Chinese and can also read English words and numbers in the same image.
Yes. Choose Simplified Chinese + English. Traditional and Simplified Chinese use separate recognition modes; the tool does not detect or translate between them.
Yes. Clear screenshots with upright, high-contrast text often work well. Small or compressed text may need a higher-resolution capture.
Blur, glare, shadows, low contrast, unusual fonts, small type, mixed directions, and compression artifacts can reduce OCR accuracy. Correct the editable result while comparing it with the source.
No. This version returns plain text from horizontal printed content. It does not rebuild columns, forms, tables, fonts, or page layout.
Convert a PDF page with PDF to JPG/PNG first. For a HEIC photo, create a JPG or PNG copy before using the resulting still image here.