On-device speech recognition, high-precision audio slicing, video frame capture, and automated romanization. Engineered specifically for rapid sentence mining.
Sentence mining is universally acknowledged as the most effective immersion technique, yet manual card authoring introduces massive friction that derails study momentum.
.apkg archive directly importable into any Anki client.Every subsystem is built natively in modern C++ with an emphasis on low overhead, offline stability, and zero external dependency friction.
No subtitle file available? The embedded speech recognition model transcribes dialogue directly on your machine without transmitting any audio data to third-party cloud infrastructure. Retains complete privacy while operating seamlessly offline.
Mark timestamps and author notes without lifting your hands from the keyboard. Jump timestamps instantly by clicking the progress timeline or stepping ±5 seconds with arrow keys.
Highlight unfamiliar terms in dialogue and trigger automated grammatical analysis. Generates concise definitions, part-of-speech categorization, and contextual example sentences.
Built-in phonetic segmentation engines for Asian languages: Pinyin with tone diacritics for Mandarin, Furigana ruby formatting for Japanese Kanji, and Romanized Hangul for Korean.
Cards stage in an adaptive two-column grid. Dynamic text expansion (1–3 lines), live audio previewing, modular media stripping, and automatic height pairing before deck export.
Different learning phases demand different cognitive prompts. Select a template below and click the card to inspect the Front and Back layout.
Template 0: Standard balanced layout. Front prompts with native dialogue and audio; Back verifies recall with contextual translation and phonetic transcriptions.
Eliminate administrative card maintenance and spend your energy where it matters: comprehending authentic native material.
Open any common media container (.mp4, .mkv, .webm). If an external .srt track exists, load it directly; otherwise rely on on-device transcription.
Press [ when speech begins and ] when the sentence concludes. Scrub the timeline directly for millisecond-level refinements.
Press S to stage the card. Export the queue to a native .apkg file and double-click to load into your Anki collection.
Core mining capabilities remain free and unencumbered. Upgrade to Pro for advanced automated cloze derivation and custom card schema design.
.apkg packages follow the official Anki SQLite media format. Once imported into your desktop Anki client, sync your collection through AnkiWeb to study on AnkiMobile (iOS) or AnkiDroid (Android).
Download the standalone portable binary for Windows 10/11. No installation required; run immediately upon extraction.