Soil carbon cycling lies at the heart of the Earth's biogeochemical cycles, significantly impacting global climate change and human health. Model simulation integrating with empirical investigations provide a robust tool for the accurate prediction of the long-term evolution and final state of soil organic carbon (SOC), and the assessment of the global SOC storage and its response to climate change. Traditional SOC models do not consider the roles of microbial turnover and mineral protection in regulating SOC, failing to reflect the real soil processes, thereby limiting their applicability in various environments or ecosystems. Modified models via the incorporations of microbial and mineral processes, have not only optimized the predictions but also deepened our understanding of the carbon cycling mechanisms. From the perspectives of model structure and process mechanisms, we first introduced two traditional models, RothC and CENTURY, as well as their limitations. We further summarized the main progresses in model improvements. Current research gaps, future research needs and outlooks were proposed to offer ideas and approaches for further studies in SOC modeling.