Abstarct: Aabstract Financial markets experience sharp price fluctuations and changes that classical models like Black–Scholes often fail to capture well. Those models are built on assumptions of normally distributed returns and constant volatility, which in practice—especially in markets with high and nonlinear volatility—do not perform adequately. This study examines and compares pure-jump models that can simulate sudden moves and extreme volatility, offering analysts better tools to model actual market behavior. In this context, Lévy processes, including the VG and NIG models, can account for skewness and kurtosis in the marginal distribution of asset returns. These features help to model price behavior and market volatility more accurately and thus can provide improved information about risks and price changes. This research evaluates the performance of these models against the Black–Scholes model, aiming to find a model that produces results closer to market realities. The study compares option pricing under three models—Black–Scholes, Variance Gamma, and Normal Inverse Gaussian—by examining several major firms listed on the Tehran Stock Exchange. Evidence shows that the mean relative error for the Variance Gamma model is smaller than that of the other two models.