Credits and licences

The open work this course is built on, and what you may do with it.

Peel Japanese is not made from nothing. The voice you hear, the stroke order you trace, and the dictionary meanings behind the vocabulary all come from people who did that work first and gave it away. They are named here, with their licences, as those licences require.

The voice#

Every spoken line in the course - the story narration, and the kana and number pronunciations - is synthesized with VOICEVOX, free Japanese text-to-speech software by Hiroshiba Kazuyuki. The voice is 四国めたん (Shikoku Metan), normal style.

Both the software and the character permit commercial use on the condition that the credit is shown where a user can find it. That credit is:

VOICEVOX:四国めたん

It sits in the footer of every page in this app. See the VOICEVOX software terms and the character voice terms. No audio file in this course is sold as a voice sample or a sound library; the narration exists only as part of the lessons.

Stroke order#

What I changed: the drawings are not used as they are. Each glyph's stroke paths were flattened from bezier curves into polylines, resampled to a fixed number of points per stroke, and normalized off KanjiVG's original coordinate space, so that a stroke drawn with a finger can be compared against them on your device without a network call. The result is a derivative of KanjiVG and is offered under the same licence - see what you may reuse.

Dictionary data#

Kanji meanings and readings come from KANJIDIC2. Vocabulary glosses derive from JMdict. Both files are the property of the Electronic Dictionary Research and Development Group (EDRDG) and are used under CC BY-SA 4.0, per the EDRDG licence. Project pages: JMdict, KANJIDIC.

What I changed: readings were transliterated into Hepburn romaji, glosses were trimmed to a length that fits a flashcard, and entries have been edited by hand where the machine-derived wording was wrong or misleading for a beginner. The data reached this app through two intermediate open datasets: open-anki-jlpt-decks and kanji-data, both MIT-licensed, whose own licences cover their code and not EDRDG's data underneath it.

The dictionary snapshot in use was taken on August 8, 2026. The EDRDG licence asks that users keep their copy reasonably current, so this is refreshed rather than frozen.

JLPT levels#

Vocabulary and kanji are ordered by JLPT level, N5 first. The JLPT is administered by the Japan Foundation and Japan Educational Exchanges and Services, who do not publish official vocabulary or kanji lists. The levels used here follow the widely used unofficial lists compiled by Jonathan Waller at tanos.co.uk. They are a community reconstruction and a good study guide, not the real exam specification. Peel Japanese is not affiliated with, endorsed by, or connected to the organisations that run the JLPT.

What you may reuse#

Some of the data in this app is open, and buying the course does not close it. Specifically:

The stroke templates are a derivative of KanjiVG and are licensed CC BY-SA 3.0.
The imported dictionary entries - the machine-derived kanji meanings, readings and vocabulary glosses - are a derivative of KANJIDIC2 and JMdict and are licensed CC BY-SA 4.0.

You may copy, share and adapt both, including commercially, as long as you credit the original authors and pass them on under the same licence. Nothing in the Terms of Service is meant to take that away, and if the two ever appear to conflict, the open licence wins for that data.

The rest is mine and is not open: the stories and their translations, the lesson design, the app itself, the generated audio, and every hand-written entry that did not come from the sources above.

Getting in touch#

If you maintain one of the works above and something here is wrong, credited badly, or missing, email me at [email protected] and I will fix it.