Open data · CC BY 4.0 · forever
The dataset belongs to everyone.
Every approved contribution is here — dialect-labelled, credited, and licensed CC BY 4.0. Build translation models, speech tech, dictionaries, keyboards. The one obligation: credit the contributors named in each record.
- 30
- Languages
- 103
- Dialects & varieties
- 2
- Approved texts
- 1
- Approved voice clips
Parallel text
Stories, proverbs, songs and essays with their English, one record per contribution, dialect-labelled.
Transcribed speech
Voice clips with their transcripts and public audio URLs — durations, prompts, speaker credits included.
Attribution
Cite as: Afrolingua Hub community dataset, CC BY 4.0, afrolinguahub.com.ng — and preserve the per-record contributor field in derived work. Each record stores the consent version its contributor accepted; the exact consent text is archived and never changes retroactively.