The viral genome of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), particularly its cell-binding spike protein gene, has undergone rapid evolution during the coronavirus disease 2019 (COVID-19) pandemic. Variants including Omicron BA.1 and Omicron BA.2 now seriously threaten the efficacy of therapeutic monoclonal antibodies and vaccines that target the spike protein. Viral evolution over a much longer timescale has generated a wide range of genetically distinct sarbecoviruses in animal populations, including the pandemic viruses SARS-CoV-2 and SARS-CoV-1. The genetic diversity and widespread zoonotic potential of this group complicates current attempts to develop drugs in preparation for the next sarbecovirus pandemic. Receptor-based decoy inhibitors can target a wide range of viral strains with a common receptor and may have intrinsic resistance to escape mutant generation and antigenic drift. We previously generated an affinity-matured decoy inhibitor based on the receptor target of the SARS-CoV-2 spike protein, angiotensin-converting enzyme 2 (ACE2), and deployed it in a recombinant adeno-associated virus vector (rAAV) for intranasal delivery and passive prophylaxis against COVID-19. Here, we demonstrate the exceptional binding and neutralizing potency of this ACE2 decoy against SARS-CoV-2 variants including Omicron BA.1 and Omicron BA.2. Tight decoy binding tracks with human ACE2 binding of viral spike receptor-binding domains across diverse clades of coronaviruses. Furthermore, in a coronavirus that cannot bind human ACE2, a variant that acquired human ACE2 binding was bound by the decoy with nanomolar affinity. Considering these results, we discuss a strategy of decoy-based treatment and passive protection to mitigate the ongoing COVID-19 pandemic and future airway virus threats.