Multilingual Cultural Q&A Benchmark Across 16 African Languages

Afri-MCQA is a multimodal cultural question-answering benchmark. It contains 8,000 Q&A pairs across 16 African languages from 13 countries, created by native speakers. The dataset was published by Atnafu and last updated in January 2026.

Use Cases

• Train multimodal models for visual question answering using the provided Q&A pairs and associated images.
• Benchmark speech-to-text models on the parallel English-African language speech data.
• Evaluate cultural bias in LLMs using native-speaker-created questions about African contexts.
• Develop cross-lingual transfer learning techniques using the parallel text Q&A pairs across 16 languages.

Strengths

• 8,000 Q&A pairs provide a substantial benchmark size.
• Covers 16 African languages from 13 countries for broad linguistic diversity.
• Data is entirely created by native speakers, ensuring cultural and linguistic authenticity.

Limitations

• Unknown row count per language may indicate potential class imbalance.
• The dataset's specific temporal and geographic coverage within countries is not detailed.
• Sample data and file formats are unavailable for immediate inspection.

Provenance

Source

Atnafu on Hugging Face.

Collection Method

Q&A pairs were entirely created by native speakers.

Time Range

null

Freshness

Last updated in January 2026.

Geography

13 African countries.