How to use it
- Choose a scanned PDF and set pages, OCR language, and DPI.
- On first use, let the browser download and cache the language model.
- Download the searchable PDF and optional TXT after recognition.
Helpful tips
- 150 DPI balances speed, memory, and recognition quality.
- Process up to 40 pages at a time; batches reduce out-of-memory risk for long documents.
- Clear, correctly oriented scans generally produce better OCR.
Frequently asked questions
Which languages are supported?＋
English, Traditional Chinese, Simplified Chinese, Japanese, and Korean models are currently available.
Is the document sent to an OCR service?＋
No. The browser only downloads a public language model; page images and recognized content stay in the local worker.