Belebele
Reading comprehension, Wolof
Official site consultedAI: benchmarks and research
Last verified: 1 October 2026
Belebele is a reading comprehension dataset covering 96 languages, usable for question answering and classification. Its size falls between 100,000 and 1 million rows. Wolof does not appear in the visible content of the page. CC-BY-SA-4.0 license.
Services
- Reading comprehension questions
- 96 languages
Access requirements
Free download on Hugging Face, CC-BY-SA-4.0 license.
Sources
- Belebele (new tab), verified on 1 October 2026Official site consulted