In this session we will present and discuss the potential of High Resolution Dynamic Spatial Models to support resilience building in cities. Integrating machine learning, geo-computation, spatial modelling and data visualisation, our team has developed a flexible modelling framework to simulate future urban scenarios according to different development pathways. The model outputs -10 m resolution grid- allow users to visually explore how their city ‘could look like’ under different trajectories, assess possible trade-offs between alternative futures and quantify potential impacts of extreme events and climate related hazards, such as coastal flooding or heat waves.We have applied our model to cities such as New York (USA), San Juan (Puerto Rico), Valdivia (Chile) and Hermosillo (Mexico), co-developing the scenarios through local collaboration with practitioners and stakeholders. In this session we will i) present the results for the different cities, ii) discuss the challenges faced on the process -e.g. data limitations, model customisation, transforming visions to algorithms, etc.- and iii) ignite a debate on how these models could contribute to foster ‘resilience thinking’ in cities.