The model can quickly search documents, whether they are text-based or include images, diagrams, graphs, tables, code, ...