Digital Twins are becoming fundamental tools to monitor the status of entities, predict their future evolutions, and simulate alternative scenarios to understand the impact of possible changes. More recently, Digital Twin solutions have been applied in the context of Smart Cities. Thanks to the large deployment of sensors, together with the increasing amount of information available for municipalities and government organizations, it is possible to build wide virtual reproductions of urban environments including structural data and real-time information that can undoubtfully help city councils and decision makers to face future challenges in the urban development and improve the citizen quality of life. In this paper, the Snap4City Smart City Digital Twin framework is presented, which is capable to respond to requirements identified in recent literature and by international forums. The proposed architecture provides an integrated solution for data gathering, indexing, computing, and information distribution, therefore realizing a continuously updated digital twin of the urban environment at global and local scales. It addresses 3D building models, road networks, Internet of Things entities, point of interests, paths, as well as results from analytical processes for traffic density reconstruction, pollutant dispersion, predictions, and what-if analysis with dynamic and constrained routing, all integrated into a freely accessible interactive 3D web interface, enabling stakeholder and citizens participation to the city decision processes. As case of study, the digital twin of the city of Florence (Italy) is presented. The solution is released on top of the Snap4City platform as open-source and made available through our GitHub repository (https://github.com/disit) and as Docker compose.