High-flow nasal cannula (HFNC) is a widely used noninvasive respiratory support technique for patients with acute respiratory failure (ARF). It provides heated and humidified oxygen at high flow rates, improving oxygenation, reducing respiratory effort, and enhancing patient comfort. International guidelines recommend HFNC over conventional oxygen therapy in hypoxemic ARF. However, there is significant variability in clinical practice regarding HFNC discontinuation, and no standardized weaning criteria currently exist. Prolonged HFNC use may increase hospital stay and healthcare costs, while premature discontinuation may lead to respiratory deterioration and the need for further ventilatory support. Previous studies suggest that successful HFNC weaning may be predicted by a Fraction of inspired oxygen (FiO₂) ≤40% and a Respiratory rate-Oxygenation index (ROX index) ≥9.2. The ROX index is calculated as the ratio of peripheral oxygen saturation (SpO₂) divided by fraction of inspired oxygen (FiO₂) to respiratory rate. The primary objective of this study is to compare a standardized HFNC weaning strategy based on ROX index and FiO₂ thresholds with usual clinical practice based on physician judgment. The primary outcome is weaning failure at the first attempt, defined as the need for HFNC reinstitution, noninvasive or invasive mechanical ventilation, or death within 48 hours after discontinuation.