<p>Serverless computing abstracts away infrastructure management, enabling developers to focus solely on event-driven function code. This model offers inherent dynamic scalability and a fine-grained cost model, but its pricing mechanisms remain a primary challenge for adoption, characterized by complexity and a lack of transparency. Unlike traditional computing models in which users are responsible for managing resources and servers, in serverless computing, users focus only on developing application logic, and the infrastructure is handled in a fully automated manner by the service provider. This model provides many advantages, including dynamic scalability, lower cost, and higher speed. However, one of the main challenges in this area is the pricing issue. Due to significant differences in pricing models, diversity of provider policies, and lack of transparency in cost structures, users are confused in predicting costs. To address this challenge, this paper conducts a systematic review of the research literature and presents a structured taxonomy for classifying pricing approaches. In this taxonomy, four main approaches are introduced: ML-based, game-theoretic-based, heuristic-based, and model-based pricing. These approaches are analyzed and compared in terms of implementation features, pricing metrics, and application areas. The paper also reviews the open issues raised in this field and provides solutions to address them. Finally, future research directions are introduced to improve the accuracy, transparency, and efficiency of the pricing process in serverless computing environments.</p>