ScanRead.ai — free OCR for handwriting, tables, and 100+ languages (Pro from $10/mo)

Hey PH! Sharing a small project we've been building: ScanRead.ai (https://scanread.ai) — a free OCR tool for extracting editable text from images, PDFs, screenshots, and handwritten notes.

What makes it a bit different from generic OCR:

• Built on PP-OCRv5 + PaddleOCR-VL (2M params) — benchmarks competitively against Google Vision and AWS Textract, especially on CJK languages

• 22 specialized tools — handwriting (including cursive), tables, math → LaTeX, receipts, screenshots, Arabic / Japanese / Chinese

• Free tier: 20 pages/day, no signup needed for casual use

• Pro: $10/mo (annual) for 3,000 pages, batch up to 20 files, .txt/.md/.docx export

• Supports JPG, PNG, PDF, WebP

Use cases we keep hearing: digitizing handwritten notes, making scanned PDFs searchable, OCR'ing receipts for expense tracking, math equations into LaTeX.

Would love any feedback or ideas for tools we should add to the suite. What OCR pain points are you running into in 2026?