Turn images into editable text.
Read printed English or Simplified Chinese from screenshots and scanned pictures. Choose your layout, correct the result and take the text with you. Free, without uploading your image.
On your device · No upload or signup. Images and text stay in this tab’s memory until you clear or leave. The recognition engine and language models download from this site; the first run may take a moment. Download your text before closing.
A clearer image gives better text.
Crop away unrelated content, keep text upright and use a sharp image. Choose “One block of text” for a paragraph, or “Scattered screenshot text” for separate labels. If the result looks wrong, check the language and rotation, then try again.
Can I extract text from a scanned PDF?
This workspace accepts JPG and PNG, not PDF files. First use PDF to images, then choose one exported page here. There is no searchable PDF output, batch OCR or spreadsheet reconstruction in this version.
What stays on my device?
The image and recognized text are processed in browser memory, not uploaded or saved to our database or browser storage. The browser may cache public engine and language-model downloads. Clear or leave the page to discard the workspace; originals, copied text and your downloads remain wherever you saved them.
What are the limits?
Printed English and Simplified Chinese only in this version. Handwriting, blurred photos, complex tables and reading order may be inaccurate. We do not promise perfect accuracy. Recognition stops after two minutes; smaller, focused images work better on phones. TXT exports contain your current edited text, not the original layout.