The conversion between low-poly and high-poly models is one of the key tasks in 3D modelling, as it determines the balance between image quality and performance. In modern industries such as video games, animation, and visualisation, high-poly models provide high detail but require significant computing resources. Low-poly models, on the other hand, optimise performance without significant loss of image quality by using techniques such as normal map burning, Ambient Occlusion and other detail maps. However, a fast and high-quality conversion between these two formats is a complex process, accompanied by problems with shading, topology, and correct map computation. This article dis cusses the main difficulties associated with converting low-poly and high-poly models, as well as methods to achieve the best results.