Apps by Satoru Nakamura
KotenOCR app icon

Reference

KotenOCR

Satoru Nakamura

Free download 3.00 · 1 storefront · 2 ratings In-app purchases unconfirmed
Open on Apple

About this app

An OCR app that reads classical Japanese cursive script (kuzushiji) using AI. Features: - Just take a photo to convert kuzushiji to text - Powered by the National Diet Library's AI model (NDL Koten OCR-Lite) - All processing runs on-device (no internet connection required) - Copy and share recognized text How to Use: 1. Capture a page with the camera or select from your photo library 2. AI automatically detects text regions and recognizes kuzushiji 3. Tap results to review, copy and use Supports over 7,000 kanji, hiragana, and katakana characters. Optimized for Edo-period woodblock prints and manuscripts. Licenses: - OCR Model: NDL Koten OCR-Lite (National Diet Library, CC-BY-4.0) - Inference Engine: ONNX Runtime (Microsoft, MIT License) Source code: https://github.com/nakamura196/koten-ocr-ios

Latest release

What’s New

Version 1.3.5 Mar 27, 2026

Improved local AI status messages. Now shows specific guidance when Apple Intelligence is not enabled or model is still downloading.

Audience growth

Ratings growth

Steady
7-day gain +0
Ratings / day 0.0
30-day gain +0
US rating-count history
2 ratings
2 Aug 19, 2026 · 2 ratings Aug 19, 2026

Baseline captured. Growth appears after the crawler observes a changed rating count.

Momentum comes from crawler snapshots for the tracked storefront. “Viral growth signal” means unusually fast recent rating gains; it is an indicator, not download data.

Price tracking

Price intelligence

1 captured event
Current Free
Historical low Free
Paid → free
Price history
Current Free
Free Aug 19, 2026 · Free Aug 19, 2026

First snapshot captured. The trend line starts after the next crawler price observation.

  1. First observed Aug 19, 2026 · 10:17 AM Free

Screenshots

KotenOCR iPhone screenshotKotenOCR iPhone screenshotKotenOCR iPhone screenshot KotenOCR iPad screenshotKotenOCR iPad screenshotKotenOCR iPad screenshot