Winter climate change has a great influence on climate change in Arctic ecosystems. Furthermore, during the dormant phase, soil greenhouse gas (GHG) fluxes can increase to noteworthy levels. GHG fluxes are thoroughly investigated, but soil organic carbon (SOC) pools are not due to modeling issues. Thus, the carbon (C) stabilization potential of the Arctic soils remained poorly assessed. We applied the two-pool model with feedback on the time-series data of CO2 fluxes derived from the long-term (180 days) laboratory incubation experiment at low (+ 4 °C) temperatures to assess the C kinetic processes, pools, and their decay rates. CO2 flux data (soil microbial basal respiration measurements) measured on four typical permafrost soils of the tundra ecosystems (Siberia) were preprocessed (bootstrapping), and a cumulative C emission was calculated for each soil type. The model outputs were C pool sizes and their breakdown rates, initial proportions of C in fast and slow pools, mean system time, and transit time. We revealed significant differences among soil types in the slow C pools and their decomposition rates. As a result, soils that respire more C and have high enzyme activity translate a minor amount of C to the microbial biomass, thus stabilizing less C in the soil, and, vice versa, soils characterized by low soil CO2 emissions and enzyme activity can stabilize C in SOC via, e.g., the growth of the microbial biomass. In summary, our findings indicate that the two-pool model with feedback may be useful for evaluating permafrost soils’ capacity to fix C and for mitigating soil GHG emissions, which are being intensified by the continuous climatic change occurring in high latitudes.