Predicting the dynamic responses of planing hulls in real sea conditions is important for identifying how basic design factors influence their seakeeping performance. Hence, there is a pressing need to provide high-fidelity models for predicting the motions of these hulls in random waves, representing actual seas. In this article, a computational-based model for solving viscous fluid flow around the vessel is built to address this problem. Three different planing hulls, denoted as C, C1, and C2, each distinguished by the number of steps incorporated on their bottom surfaces (1 and 2 indicating the respective step count, with case C being the stepless hull), are modeled in a Computational Fluid Dynamics (CFD) tank, allowing for analysis of the effects of steps on dynamic responses of a planing surface operating in random waves. CFD data is compared against those collected in towing tank tests, revealing a satisfactory level of accuracy. Extreme value and gamma distributions are shown to give probabilities of maxima/minima of displacements and vertical acceleration at the center of gravity (CG) for all three hulls. It is shown that the stepless boat may be exposed to lower vertical acceleration at an early planing speed, but at higher planing speeds, a double-stepped design mitigates the vertical acceleration. Nevertheless, the double-stepped hull would experience more significant extreme heave responses across all speeds and may be exposed to less significant extreme pitch responses during the ride at the highest speed compared to the stepless and one-stepped hulls. The skewness of heave and pitch is evaluated, and it is found that the heave response tends to skew toward positive values (upward). This skewness becomes more noticeable with increasing speed but remains insensitive to wave steepness. Additionally, the pitch response at lower planing speeds shows a partial skew towards negative values (bow-down), but eventually, they may also be partially skewed towards positive values at higher speeds. Moreover, a correlation is observed between the kurtosis of responses of different hulls and the occurrence of the 1/100 highest responses, indicating that a kurtosis greater than 3.0 would result in more extreme responses. Overall, this analysis offers practical insights into planing hull behavior in actual sea conditions from a CFD model perspective, highlighting the potential of CFD in simulating this complex problem.