Docking-based virtual screening is a structure-based drug design approach widely used by the scientific community to assist drug discovery in searching for new lead compounds against relevant therapeutic targets [1], [2], [3], [4]. The primary purpose of this methodology is to predict receptor-ligand binding modes and identify potential hits by docking a dataset of compounds against the three-dimensional structure of one or more receptors of interest [5], [6], [7]. The DockThor program, developed by our research group GMMSB/LNCC, has obtained promising results in comparative studies compared to other well-established docking programs for predicting experimental binding modes and binding affinities considering several molecular targets and chemical classes of ligands, including highly flexible peptides [8], [9]. Given the emerging importance of docking-based virtual screening in drug discovery and the promising accuracy of our recently developed methodologies, in this work, we describe the DockThor-VS as a free online platform provided by the Brazilian high-performance platform SDumont. This project is a significant effort towards facilitating greater access and efficiency of this technique among academics and researchers. In 2020, the validated accuracy of the DockThor program for binding mode and affinity predictions encouraged us to improve the DockThor-VS portal to accelerate the fight against the COVID-19 pandemic as a free and reliable tool for virtual screening. As a result of our efforts during this period, we provided the structures of potential SARS-CoV-2 therapeutic targets, including clinically relevant non-synonym variations and distinct conformations representative of the flexibility of each receptor ready to dock, providing a free large-scale virtual screening service to the scientific community to help fight against COVID-19 and further emergency situations [10]. More recently, we also included in the DockThor-VS web server datasets of compounds, including an updated database of the FDA-approved drugs with protonation and tautomeric states within different pH ranges to be used in virtual screening experiments. These datasets constitute a key asset, enabling virtual repositioning studies within the scientific community.