Nigerian artificial intelligence startup Veta Origin has officially launched a localized large language model across six countries, aiming to build an AI system that truly fits the African context rather than merely adapting mainstream Western models. The model is currently open for testing by real users, with the team rapidly iterating based on feedback, aiming to make advanced AI technology more aligned with the daily needs of African users.
Current mainstream AI models globally, such as ChatGPT, Claude, and Gemini, primarily rely on Western data for training, resulting in poor performance when addressing African languages, cultural backgrounds, and regional issues. Even though some models support African languages like Swahili and Hausa, their understanding accuracy and semantic expression remain significantly below standard, primarily due to a severe lack of training data.
The founder of Veta Origin pointed out that before launching the project, the team conducted in-depth research into the actual needs of various African markets, systematically planning AI infrastructure and localization integration solutions, and assembling a multidisciplinary team with capabilities in technical development, operational implementation, and community outreach.
Unlike global giants that view African languages as an additional feature, Veta Origin has included Hausa, Igbo, Yoruba, and Swahili in its core support system from the very beginning. This move not only covers hundreds of millions of native speakers but also fills the gap in the mainstream AI ecosystem regarding African languages, truly targeting a large, overlooked market.
As African countries actively seek to participate in the global AI process, the necessity for localized AI is becoming increasingly evident. Models reliant on Western training often fail to accurately address local issues in education, healthcare, governance, or business scenarios. Language models led by African teams and built on African data are expected to provide more precise and practical intelligent services for a population of 1.4 billion, unlocking growth potential that existing technologies have overlooked.
Although Veta Origin is still in its early development stages and faces a resource gap compared to giants like OpenAI and Google, its competitiveness remains to be tested. However, the simultaneous launch across six countries has already demonstrated its firm commitment to building an AI infrastructure exclusive to Africa, rather than passively waiting for international companies to make adaptive adjustments. Currently, the team is focusing on gathering early user feedback to continuously optimize model performance and expand functional boundaries.