PDF OCR
Extract text from scanned PDFs with OCR, right in your browser. Supports 10 languages including Arabic, Hindi, and Chinese, free, private, no uploads.
About This Tool
Turn a scanned document into text you can search, copy and edit. Each page is rendered and read by an OCR engine that runs inside your browser, and the recognised text comes back as a plain .txt file with page markers. Ten languages are supported, including Arabic, Hindi and Chinese. Your PDF is never uploaded, only the recognition model is downloaded, once.
How To Use
- Drop the scanned PDF into the upload zone or click to select it.
- Pick the language of the document and a scan quality.
- Click "Extract Text" and wait, the first run downloads the OCR engine.
- Download the .txt file with the recognised text.
How It Works
OCR needs pixels, so each PDF page is first rasterised with pdf.js at the DPI you choose. Higher DPI gives the recogniser more detail, 300 DPI noticeably improves accuracy on small print, at the cost of speed and memory.
The rendered page is then passed to Tesseract, a mature open-source OCR engine compiled to WebAssembly. It segments the image into lines and words, matches shapes against a trained model for your language, and returns the text with page markers so you can tell where each page starts.
The first run for a language downloads that language model (a few megabytes) and the browser caches it for later. Recognition itself is CPU-heavy: expect a few seconds per page, more at 300 DPI. Everything happens on your device, the document never leaves it.
Frequently Asked Questions
Why is the first run slow?
The OCR engine and the language model are downloaded the first time you use a language. After that the browser serves them from cache and later runs start immediately.
How accurate is it?
On a clean 300 DPI scan, accuracy is typically very high. Blurry photos, skewed pages, unusual fonts and handwriting are much harder, OCR does not read handwriting reliably.
Do I get a searchable PDF?
Not yet. The output is a plain .txt file containing the recognised text, page by page.
Is my document uploaded?
No. The PDF is read and recognised entirely in your browser. Only the language model is downloaded, and it contains no data about your file.
My PDF already has text, do I need OCR?
No. If text can already be selected in your reader, the PDF to Word tool will extract it faster and more accurately.