Built on generous shoulders.

Wakanai exists because people published useful work under licenses that let a small app build on it. This page is the ledger.

Open data

• Kanji metadata: from the Electronic Dictionary Research and Development Group (EDRDG) dictionary files, used under the group's licence (CC BY-SA). This data powers the kanji-readiness gating that decides when harder script is fair to show you.
• Word frequency data: from hermitdave/FrequencyWords (CC BY-SA), derived from the OpenSubtitles corpus. It helps order which words matter first.

Voice

• Toro speaks with にせ (Nise), a voice model by 魔法プログラム, published on AivisHub under ACML 1.0 and synthesized with the open-source AivisSpeech engine. Credit is optional under the license. This line is gratitude, not obligation.
• Earlier beta builds used VOICEVOX:四国めたん for word pronunciations, via the free VOICEVOX engine.

The podcast

Lessons are built around episodes of Nihongo con Teppei for Beginners by Teppei, streamed from the podcast's own published feed. Attribution also appears inside every lesson, where it belongs. Wakanai is an independent app, not affiliated with or endorsed by the podcast.