Google’s India-first farm models expand to African markets

Google said on 26 August that agricultural AI models first developed around India were providing insights in Kenya, Uganda, Ghana, Rwanda, Nigeria and Zambia. Its research update describes Agricultural Landscape Understanding and Agricultural Monitoring & Event Detection, with outputs available through free APIs and integrated into Google Earth.

The tools analyse agricultural landscapes and activity, supplying information that other organisations can combine with records and local services. Their expansion is an applied AI development beyond the consumer chatbot market.

Google points to CarbonFarm’s use of its models and Gemini for rice-farming monitoring, and Terrastack’s use of satellite information alongside land records and climate signals for farm-related decisions in India. These are examples reported by Google, not independently measured Capvise outcomes. The access route is concrete: developers can work with field-level model outputs, while the usefulness of a resulting credit or climate service depends on its wider data and implementation. The update documents that expansion.