CaptionCatch Autor: NKaouri

CaptionCatch turns the subtitles you're already watching into passive vocabulary practice.

As a word appears in the captions on YouTube or Netflix, CaptionCatch picks one useful word from the line and shows it in a small running list over the video — the word, its reading, and its meaning — then moves on. No pausing, no lookups. You just watch.

Three source languages
- Japanese — tokenized with kuromoji, graded by JLPT level
- Chinese — segmented with jieba, graded by HSK level
- English — graded by CEFR level (CEFR-J + WordNet)

Show meanings in English, German, Spanish, or French. Set your level and how often words appear.

Fully offline dictionary
Word lookup runs entirely on your device — bundled tokenizers and graded word lists. No AI, no API, no account, works with no internet. CaptionCatch reads the caption text already rendered on the page; it never touches the video stream or subtitle files.

Also included
- Stats dashboard — streaks, a 90-day activity map, level and platform breakdowns
- Export learned words to an Anki-ready CSV
- Filter which word types you see (noun / verb / adjective / adverb)
- Light / dark / system theme, six overlay positions
- A running review list of every word you've seen

Free vs Pro
Free on YouTube (up to 120 words/day), all three languages, the overlay and review list. CaptionCatch Pro (optional subscription) adds Netflix support, unlimited words, the stats dashboard, Anki export, and the word-type filter.

Privacy
Your word history, settings, and stats stay in your browser — nothing is sent anywhere. The only network request is a periodic license check to Gumroad for Pro subscribers, sending only the license key.