With the rapid development and intelligent evolution of large language models(LLMs) and the automotive industry, sound effect algorithms for vehicle play a crucial role in enhancing the acoustic experience within the cabin.In this paper,a comprehensive review of sound effect algorithms, functions, system architectures, and applications for vehicle based on the LLMs was provided. Firstly, it introduces the development history and current technological status of the in-vehicle sound effects,along with future application scenarios and trends in this field driven by LLMs. Next, it analyzes the advantages of more than 10 product functions and algorithms from three dimensions: sound reproduction, noise management, and intelligent experience.The paper further discusses challenges and near-term solutions for implementing these algorithms in vehicles.Finally, Large language models conduct data-driven learning and model training on audio systems of commercially available vehicles.This process enabled the development of a reference framework for enhancing in-vehicle acoustic systems across three critical dimensions: audio chip, amplifiers, and speaker. This review aims to serve as a reference for research on LLMs based smart car sound effects, promoting widespread adoption in the automotive field.