Influenza viruses are segmented negative-sense RNA viruses with high genetic variability, causing seasonal epidemics and pandemics. Whole-genome sequencing from clinical specimens is often hindered by low viral loads and host background, necessitating enrichment-based next-generation sequencing (NGS). In this study, we compared amplicon- and probe capture-based NGS using 26 influenza-positive clinical specimens representing A(H1N1), A(H3N2), and B/Victoria lineages, stratified by cycle threshold (Ct) (<25, 25-30, ≥30). Sequencing performance was assessed based on mapped read proportion, mean coverage depth, and segment-level genome recovery across the eight influenza gene segments. For specimens with Ct <25, amplicon-based NGS showed higher mapped read proportions, while both methods achieved near-complete genome recovery. In the 25 ≤ Ct <30 group, overall genome recovery was comparable between approaches, although subtype-specific differences were observed. For Ct ≥30 specimens, probe capture-based NGS demonstrated significantly higher mapped read proportions, greater mean coverage depth, and near-complete segment recovery for influenza A viruses, whereas amplicon-based NGS showed reduced recovery with frequent segment dropouts. Our findings provide a systematic comparison of enrichment strategies under standardized conditions and offer evidence to inform methodological selection for influenza whole-genome sequencing across diverse clinical viral load ranges.