Soils store the largest amount of carbon (C) in the biosphere, and the C pool in soil is critical to the global C balance. Numerous microbial models have been developed over the last few decades to represent microbial processes that regulate the responses of soil organic carbon (SOC) to climate change. However, the representation of microbial processes varies, and how microbial processes are incorporated into SOC models has not been well explored. Here, we reviewed 70 microbial models to characterize the microbial processes incorporated into SOC models and analyzed variations in mechanistic complexity. We revealed that (1) four processes (microbial decomposition, mineral interaction, microbial mortality, and transition between active and dormant microbial states) are commonly incorporated in microbial models, (2) ~47% of models simulate only one (i.e., microbial decomposition) and 33% simulated two microbial processes: microbial decomposition and mineral interaction, (3) more than 80% microbial models use nonlinear mathematical equations, such as forward Michaelis-Menten kinetics, to represent SOC decomposition, (4) the concept of persistence of SOC due to its intrinsic properties has been replaced by organo-mineral interaction (~39% of microbial models) that protects SOC from decomposition, and (5) various temperature and moisture modifiers and pH effects have been used to explain the environmental effect on microbial processes. Finally, we propose a roadmap for SOC model improvement. In the future, to realistically incorporate microbial processes into Earth System Models, it is imperative to identify experimental evidence on rate limitation processes and firmly ground model structure on the field and laboratory data.