This chapter addresses how Graph Neural Networks (GNNs) could mix several kinds of data to address challenges. Multimodal data (MD) is the application of data from several sources like text, photos, sensor measurements, etc. Using only one kind of data is often insufficient to completely grasp a problem. Combining several types of data helps one to identify better results. GNNs transform this data into a graph, including nodes, edges, and shared knowledge. Early fusion, late fusion, hybrid fusion, attention mechanism, and tensor fusion are just a few of the several ways data could be combined. In fields including healthcare, self-driving cars, online retail, and social media, GNNs utilizing multimodal data find application. This chapter describes how GNNs handle this kind of data and how they might tackle several kinds of problems.