Engine not loaded
This is a developer test of the OCR pipeline — not the product. The shipping product is an iPhone app powered by Apple Intelligence on iOS 26. View the iPhone mockup →
Local test harness v0.2 (Tesseract.js classical OCR). All processing happens in your browser. OCR engine (~5MB) loads on first use and caches locally. Documents never leave your device.

Input 0 files

Drop documents here

JPG, PNG, WebP, HEIC, PDF — or click to choose

    Extracted output 0 processed

    No results yet. Load the model, drop files, click "Process all".

    AI-drafted for your review. Extracted fields are first-pass and may contain errors. Verify every value before using in any real context. This harness is for accuracy testing and development — not for production document handling.