Free · Private · No upload
OCR PDF to text online
Recognize English, Spanish, French, German, Italian, Portuguese, Dutch, Arabic, Chinese or Japanese text in scanned PDFs without uploading the document.
High-quality PDF OCR that stays in your browser
Every page is rendered near 300 DPI within a safe memory limit, recognized with a quality-optimized LSTM language model and rebuilt in reading order. One worker is reused across the document and the model is cached after its first download.
- Choose a scanned PDF
Add an image-only or difficult PDF and select the document language.
- Run local OCR
Pages are rendered and recognized on your device using a reusable quality OCR worker.
- Download searchable text
Review and download the recognized content as a standard UTF-8 TXT file.
Common questions
Is the scanned PDF uploaded for OCR?
No. Page pixels and recognized text stay in your browser. Only the OCR code and selected language model are downloaded from the package CDN on first use.
Which languages are supported?
The interface offers English, Spanish, French, German, Italian, Portuguese, Dutch, Arabic, Simplified Chinese and Japanese quality models.
Why does the first OCR run take longer?
The browser must download and initialize the selected language model once. It is cached for later runs, and one worker is reused for every page in the document.