South Africa’s Vambo AI has emerged as one of the standout recipients of Meta’s Llama Impact Grant, an initiative designed to support artificial intelligence (AI) innovation across Sub-Saharan Africa. Co-founded by Chido Dzinotyiwei, Vambo AI is developing multilingual AI infrastructure that supports translation, transcription, content generation, and search across more than 60 African languages. By treating language as critical infrastructure, the start-up addresses the persistent challenge of digital inclusion in a continent marked by linguistic diversity.
The Llama Impact Grant, established in October 2023 in partnership with Data Science Africa, provides recipients with $20,000 (approximately R346,000) alongside technical mentorship, networking opportunities, and access to stakeholders in research, policy, and industry. Since its launch, the programme has attracted more than 800 applications from over 90 countries, reflecting Africa’s increasing centrality in AI development and application.
For Vambo AI, the award is more than financial support. It signals recognition of the importance of creating digital tools rooted in African contexts. In a landscape where global AI models often marginalise underrepresented languages, Vambo’s work highlights how AI can be designed to empower rather than exclude.
South Africa’s PropelMapper was also named among this year’s winners. Founded by Reghardt Adriaan Pretorius and Mark Donne, the start-up provides agricultural advisors with AI-driven tools to produce customised podcasts for farmers, generate professional advisory reports, and integrate satellite imagery for intervention alerts. This integrated approach seeks to advance food security and improve productivity for farming communities.
Beyond South Africa, the 2025 grant cycle recognised several other African innovators. In Nigeria, Taiwo Oyewole’s Radease equips patent medicine vendors with AI tools delivered through WhatsApp to simplify access to trusted health information. In Uganda, Rahman Sanya’s TeenApp delivers youth-friendly reproductive health education through digital platforms guided by responsible AI practices. In Rwanda, Isaac Manzi’s Easy Read Africa transforms complex documents into accessible text and visuals for individuals with cognitive or learning challenges.
Earlier awardees have included Farmer Chat, designed to support smallholder farmers, and Jacaranda Health’s Prompts, which provides maternal health interventions. Together with current winners, these projects demonstrate that African-led AI development is not confined to abstract research but is embedded in urgent social and economic contexts.
Meta’s most recent open-source model, Llama 3.3, is freely available for adaptation and has already been deployed in education and healthcare applications across low-resource settings. The recognition of ventures like Vambo AI underscores how African start-ups are shaping global conversations about technology by reframing AI as a tool for inclusivity, access, and sustainability.
