The emergence of resistance to antiviral drugs increasingly used to treat SARS-CoV-2 infections has been recognized as a significant threat to COVID-19 control. In addition, some SARS-CoV-2 variants of concern appear to be intrinsically resistant to several classes of these antiviral agents. Therefore, there is a critical need for rapid recognition of clinically relevant polymorphisms in SARS-CoV-2 genomes associated with significant reduction of drug activity in virus neutralization experiments. Here we present SABRes, a bioinformatic tool, which leverages on expanding public datasets of SARS-CoV-2 genomes and allows detection of drug resistance mutations in consensus genomes as well as in viral subpopulations. We have applied SABRes to detect resistance-conferring mutations in 25,197 genomes generated over the course of the SARS-CoV-2 pandemic in Australia and found 299 genomes containing resistance conferring mutations to the five antiviral therapeutics that retain effectiveness against currently circulating strains of SARS-CoV-2 – Sotrovimab, Bebtelovimab, Remdesivir, Nirmatrelvir and Molnupiravir. These genomes account for a 1.18% prevalence of resistant isolates discovered by SABRes, including 80 genomes with resistance conferring mutations found in viral subpopulations. Detection of these mutations within subpopulations are critical as these mutations can become an advantage under selective pressure and presents a significant and important step forward in our ability to monitor SARS-CoV-2 drug resistance.