Credits & sources

Salanguit’s word bank is built from hand-written lessons, open dictionary data and community contributions. Every new word goes through duplicate checks, an accuracy review and (for imported data) a second AI check before it appears in lessons.

Open data we build on

• Wiktionary, extracted by kaikki.org (Wiktextract) – licensed under CC BY-SA 4.0. Used for words and meanings.
• Tatoeba – example sentences and translations by the Tatoeba community, licensed under CC BY 2.0 FR. Used for short example sentences.
• Word-frequency lists from open subtitle and web corpora – used only to decide which words are common enough to teach, and at which CEFR level.

CEFR levels

Levels follow the Common European Framework of Reference. Levels for imported words are estimates based on how common a word is; see what the levels mean.

Found a mistake?

Please tell us from the feedback button inside the app – we fix wrong words quickly.