In recent years, the significant increase both in terms of frequency and intensity of heat waves has dramatic consequences in urban areas. The amplification of the urban heat island (that can exceed 100 C) has indeed hazardous impacts on health or biodiversity and it becomes urgent to set up early warning systems. Understanding and simulating the Urban Heat Island (UHI) is therefore of prime importance to help city planners in the conception of more resilient cities. In practice, because of the complexity of this phenomenon, no trivial model exists and number of contributions have been proposed (mostly based on physics or statistics), all of them requiring a variety of input parameters related to urban morphology, land use, material properties or meteorological conditions. In this paper, we analyze the performances of the open source Urban Weather Generator (UWG) model in the context of Toulouse, one of the biggest French cities. Morphological and land cover parameters are derived from remote sensing and performances of UWG are analyzed on various Local Climate Zones (LCZs). Simulated air temperatures showed good consistency on different LCZs with a mean bias error (MBE) ranging between ´0.510 C to 0.370 C and a root mean square error (RMSE) between 1.730 C to 2.350 C. We also show that best performances are achieved on dense urbanized areas.