Text, unlocked.
Extract English and Simplified Chinese text from images and scanned PDFs.
Your file stays on your device. On first use, the OCR engine and language models are downloaded from third-party servers; an internet connection is required.
Your document
No file selected.
Use 90 or -90 for sideways pages. Any angle from -180° to 180° is supported. Applied to every PDF page.
Automatic straightening works best for printed text tilted up to about 30°. It may not detect sparse or complex layouts.
Up to 20 PDF pages. Clear, upright printed text works best.
Extracted text
Review the result for errors. Layout, tables and handwriting may not be preserved.
How it works
Select a document and its language, then choose Extract text. The result stays in this tab until you leave or replace the file.
PDFs are rendered page by page. Password-protected PDFs are not supported. Text recognition is not guaranteed to reproduce the original formatting.