TerraMind is IBM’s first generative, multimodal foundation model for Earth observation (EO). The model takes geospatial data understanding to new levels, introduces new capabilities such as Thinking-in-Modalities (TiM), and significantly outperforms existing models across community-standard benchmarks. In this session, we will explore the concept of foundation models and how TerraMind functions. Then, using the TerraTorch foundation model toolkit, we will create multimodal generations and fine-tune TerraMind for a downstream use case in an interactive session.