ZZZ Let's Read!

ZZZ Let's Read! — On-device OCR for Japanese (vertical, horizontal, classical,
kuzushiji) and English

Point your camera at a book or a receipt, and get text in an instant. Every
byte of OCR runs locally on your device — captured images and recognized text
never leave your phone. Ideal for privacy-conscious users and offline
environments.


◆ SUPPORTED OCR LANGUAGES AND SCRIPTS ◆

This App recognizes only the following:
• Japanese — vertical (tategaki) and horizontal (yokogaki) writing, modern
Japanese, classical/pre-modern kanji, and kuzushiji (cursive handwriting).
7,141 characters of the NDLmoji charset.
• English — basic Latin alphabet (A–Z / a–z), digits, and common punctuation.

Other writing systems are NOT recognized: Simplified/Traditional Chinese,
Korean (Hangul), Cyrillic, Arabic, etc. Accented Latin characters (é, ñ, ü, ç,
ã, …) are also not in the OCR charset. The app interface is available in 6
languages, but the OCR engine itself supports only Japanese and English.


◆ FEATURES ◆

Document Mode — Books, Manuscripts, Classical Japanese
• Powered by the National Diet Library's NDLOCR-Lite
• 7,141-character NDLmoji charset including pre-modern Japanese kanji
• "Detect" outlines all text regions — tap to OCR only what you need
• "Read all" runs OCR on the entire page in one go
• Copy or share recognized text
• Pinch-to-zoom on the captured snapshot

Receipt Mode — Automated Bookkeeping
• Auto-extract amount and date from any receipt
• Multi-currency: JPY, USD, GBP, EUR, and more
• Date formats: Gregorian, Reiwa/Heisei, abbreviated wareki, English month
names (e.g. 16AUG2015)
• Auto-detects DD/MM vs MM/DD by device locale
• Handles European decimal comma (15,65 €) correctly
• Bill splitter — divides the amount by N people
• Share CSV via Mail / Drive / messenger, or clear with one tap


◆ SIX INTERFACE LANGUAGES ◆

Japanese, English, Chinese, Korean, Spanish, Portuguese.
Unsupported locales fall back to English automatically.


◆ PRIVACY BY DESIGN ◆

• Fully offline OCR — no network required for recognition
• Captured photos are discarded after OCR; nothing is sent to a server
• Only the camera and one CSV file on internal storage are accessed
• Network is used solely for ad delivery (Google AdMob)


◆ WHO IT'S FOR ◆

• Students and researchers studying classical Japanese
• Writers and editors quoting from physical books
• Freelancers, sole proprietors, and households tracking receipts
• Travelers logging expenses across multiple countries
• Anyone handling confidential documents or working offline


◆ REQUIREMENTS ◆

• iOS 13.0 or higher
• ~200 MB free storage
• Rear-facing camera
• Internet not required for OCR (used only for ads)


◆ OPEN-SOURCE CREDITS ◆

The OCR engine in this App is built on these open-source assets, redistributed
under their respective licenses.
• National Diet Library NDLOCR-Lite — CC BY 4.0
• PARSeq scene-text recognition model — Apache License 2.0
• PaddlePaddle PP-OCRv4 detection model — Apache License 2.0
The native iOS port (Swift + ONNX Runtime) is the work of this App's
developer.


◆ MONTHLY SUBSCRIPTION (REMOVE ADS) ◆

• The "Remove Ads" monthly plan hides all ads in the app
• Payment is charged to your Apple ID account
• The subscription renews automatically unless auto-renew is turned off at
least 24 hours before the end of the current period
• Your account is charged for renewal within 24 hours prior to the end of the
current period
• You can manage or cancel anytime in iOS Settings > Apple ID > Subscriptions


◆ TERMS OF USE (EULA) ◆

This App is provided under the following End User License Agreement (EULA).
Terms of Use: https://smolt-stage.com/home/privacy/zzz-ocr-terms.html

◆ PRIVACY POLICY ◆
This App handles information in accordance with its privacy policy. For
details, see "About" inside the App or the link in this store listing.