<title>Abstract</title> Earthquake networks are a representation of seismicity which has brought the use of graph theory conceptual tools to seismology. This research delves into the topologies and complexities of \textit{cells-as-nodes} earthquake networks and tries to elucidate the effect of changing the cell size to create such networks in the resultant topology. In essence, the cell size provides a multiscale microscope that allows to establish the topological properties of the network at different levels. Focusing on the seismicity of Colombia, this research compares earthquake networks with simulated network models, such as Erdős-Renyi, Watts-Strogatz, and Barabási-Albert, and looks into different metrics such as entropy, Complexity, or small-worldness. We find that smaller cell sizes enhance heterogeneity and complexity in earthquake networks; that regardless of cell size, exhibit small-world and scale-free characteristics, with noteworthy variations as we change the cell size.