AI4Wolof
Gemini, Claude, ChatGPT comparison in Wolof
Official site consultedAI: benchmarks and research
Last verified: 1 October 2026
Repository evaluating 13 large language models (Gemini, Claude, GPT-4, DeepSeek, Mistral, etc.) on Wolof: translation, generation, comprehension and grammatical analysis. In translation, Gemini 2.0 Flash scores highest (0.685), ahead of Claude 3.5 Sonnet (0.663). MIT license.
Services
- Ranking of 13 models by score
- Corpus of conversations, news, literary and technical texts
Access requirements
Public GitHub repository under MIT license.
Sources
- AI4Wolof (new tab), verified on 1 October 2026Official site consulted