What it does
Recognize printed text from scanned pages with Tesseract in a local worker. Output is editable TXT; review and correct recognition errors.
When to use it
Use it for your own documents or documents you are authorized to edit. Keep the original and verify the output.
Steps
- Load a PDF and select up to five pages.
- Choose Spanish, English or both and 108/144 DPI. Local engine and model downloads start only when processing.
- Follow per-page progress or cancel. Review the text before copying it or downloading TXT.
Example
Readable scan: El niño pidió información. Hello world.
Editable text containing those words, subject to errors depending on image quality, resolution and language.
Common mistakes
PDF 5 MiB/50 pages; select up to 5. Raster 8 MP/page and 24 MP/job; OCR timeout 120 seconds.
No searchable PDF, table reconstruction or handwriting guarantee. Blank pages may return empty text. Accuracy depends on the image.
Limitations
PDF 5 MiB/50 pages; select up to 5. Raster 8 MP/page and 24 MP/job; OCR timeout 120 seconds.
No searchable PDF, table reconstruction or handwriting guarantee. Blank pages may return empty text. Accuracy depends on the image.
FAQ
Are documents uploaded?
Processing stays in your browser. Visiting the site downloads its resources, including OCR models when requested.
Should I check the result?
Yes. Keep the original and inspect the copy or recognized text before relying on it.