This research presents a comprehensive framework for creating digital twins of buildings by combining cloud mapping platforms, multi-agent LLMs, and advanced 3D modeling techniques.

• Leverages Google Maps Platform APIs to access geospatial data
• Utilizes ChatGPT-4o and Deepseek models for intelligent data analysis
• Implements Gaussian Splatting for high-quality 3D mesh extraction
• Creates detailed virtual replicas for enhanced decision-making

This integration enables engineers and urban planners to optimize infrastructure management through accurate, data-driven virtual representations, ultimately improving design, maintenance, and operational efficiency in the built environment.

Digital Twin Buildings: 3D Modeling, GIS Integration, and Visual Descriptions Using Gaussian Splatting, ChatGPT/Deepseek, and Google Maps Platform