Where Are We? Evaluating LLM Performance on African Languages

TL;DR

This paper evaluates the performance of large language models on African languages using a new benchmark, revealing disparities caused by data inequities and emphasizing the need for inclusive data practices and policy reforms.

Contribution

It introduces Sahara, a comprehensive benchmark for African languages, and provides an empirical assessment of LLMs, highlighting the impact of data disparities and proposing actionable policy recommendations.

Findings

Few African languages perform well due to data scarcity.

Most Indigenous languages are marginalized in current LLMs.

Policy reforms can improve linguistic inclusivity in AI.

Abstract

Africa's rich linguistic heritage remains underrepresented in NLP, largely due to historical policies that favor foreign languages and create significant data inequities. In this paper, we integrate theoretical insights on Africa's language landscape with an empirical evaluation using Sahara - a comprehensive benchmark curated from large-scale, publicly accessible datasets capturing the continent's linguistic diversity. By systematically assessing the performance of leading large language models (LLMs) on Sahara, we demonstrate how policy-induced data variations directly impact model effectiveness across African languages. Our findings reveal that while a few languages perform reasonably well, many Indigenous languages remain marginalized due to sparse data. Leveraging these insights, we offer actionable recommendations for policy reforms and inclusive data practices. Overall, our work underscores the urgent need for a dual approach - combining theoretical understanding with empirical evaluation - to foster linguistic diversity in AI for African communities.