Nowadays, manufacturers often sell different products (models of a product) through both online and offline channels. To face challenges such as legislation or competition, manufacturers often need to provide warranties for their products. On the one hand, customers are more attracted to goods covered with a warranty; on the other hand, warranties generate additional costs for manufacturers. This study investigates a case that a manufacturer sells through online and offline channels and offers a two-dimensional warranty policy including a warranty-age and a warranty-usage package for sold products through online channel. Because different models share several components, failures of all manufacturer’s portfolio products are statistically dependent. A model to optimize both the pricing and warranty policies is proposed. Given that the claim rates for warranties are stochastic, the value-at-risk approach is implemented to solve the optimization problem. Furthermore, the covariance between warranty claims associated with different products is computed through a copula. The findings indicate the importance of considering the covariance among different models claims when optimal warranty policy is offered to customers along with the proper pricing strategy in an online channel.