Recognize text from rendered PDF pages using local Tesseract OCR.
Processing: file or text contents are handled in your browser. Libraries and models may download over the network. Optional page analytics is controlled through Privacy choices. Privacy details
OCR language data (~10 MB) is downloaded on first use. Processing is done entirely in your browser.
Select a clear scan and its language. First use downloads the OCR engine and language data.
Run OCR. Proofread names, numbers and punctuation against the image.
Accuracy and time depend on scan quality, language and device. No measured accuracy is claimed.
PDF highlights may not align with words. Long lines, page overflow or unsupported font characters stop PDF output; use TXT to retain the recognized text.
Recognize text from rendered PDF pages using local Tesseract OCR.
Input: A PDF, up to 50 MB, and the selected OCR language. Output: TXT or an image-based PDF with an approximate invisible text layer.
Keep your original. No fixed processing time, guaranteed output size or universal browser support is claimed.
Optional Google Analytics helps us understand page visits and tool results. File contents, filenames, passwords and text inputs are excluded from our tool events. Tools work with either choice. Privacy details