This article evaluates the accuracy of 3D models made from point clouds obtained from photogrammetry. Photographs were taken from ground level and using a drone, and data processing was performed in 3DF Zephyr. The models were compared with the actual dimensions of the buildings. Four different building objects with varying degrees of complexity were analysed. The aim of the research is to analyse the conditions for taking photographs and how they are transformed into a point cloud, and to see how and whether the complexity of the shape of the facade affects the accuracy of the 3D model made from the point cloud. The inaccuracy of the point cloud in the form of point spread for all analysed cases was 1.8±0.4 cm on average. The largest measurement error was found in the case of a multi-storey building. Despite the presented inaccuracies, it was considered advantageous to use the point cloud obtained through photogrammetry in the inventory. No difference was observed in the accuracy of the model depending on the complexity of the building. Recommendations were made regarding the conditions for taking photographs.