The rapid diagnostics tests for SARS-CoV-2 antigen vary in their sensitivities, and moreover, genomic mutations may further affect the performance of the assays. We aimed to evaluate the analytical performance of an automated antigen assay and compare its sensitivity in Delta- and Omicron-variant positive clinical samples. The analytical performance of an automated mariPOC SARS-CoV-2 antigen test was evaluated on a population of community-dwelling subjects with mild respiratory symptoms or being asymptomatic investigated by the RT-qPCR Allplex™ SARS-CoV-2 assay. The sensitivity and specificity of the antigen test were evaluated on prospective 621 nasopharyngeal swabs along with oropharyngeal swabs. The sensitivity regarding variants determined by the Allplex™ SARS-CoV-2 Variant assays was analysed in additional, retrospective 158 Delta and 59 Omicron samples. The overall sensitivity of the antigen test in prospective samples was 77.9% (113/145; 95% confidence interval [CI] 70.3-84.4) with the specificity of 99.8% (95% CI 98.8-100). Regarding the variant, the sensitivity was higher in Omicron-variant samples, 93.2% (55/59; 95% CI 83.5-98.1), compared to Delta-variant samples, 71.5% (113/158; 95% CI 63.8-78.4; p = .001). In community-dwelling subjects with mild respiratory symptoms or being asymptomatic, the automated mariPOC SARS-CoV-2 antigen test showed high sensitivity over 98.0% in subgroup samples with cycle threshold (Ct) values < 25. Regarding the variant, the antigen test sensitivity was higher in the Omicron-variant samples compared to the Delta-variant samples. The analytical performance of the antigen test can differ between the SARS-CoV-2 variants, and a re-evaluation should be performed for new circulating lineages.