Soot generation is a challenging issue in high-temperature biomass gasification, which reduces the biomass conversion rate and leads to contamination of the reactor. To provide new means and insights to optimize gasification processes, the soot generation during biomass gasification in a cyclone reactor is studied here by establishing a novel biomass gasification and soot formation model to improve the accuracy attainable in numerical predictions of spatio-temporal soot evolution. The new method is validated by comparing it with gasification experiments in two reactor configurations. A good performance in capturing the overall soot generation and light gas yield of the current model is obtained in the simulations of an entrained flow reactor compared with experimental data. Besides, the biomass gasification behavior in this entrained flow reactor is systematically studied by reviewing the tar, precursor, and soot mass fraction evolution in the reactor under different steam/carbon ratios, gasification temperatures, and air excess ratios with the new model. Furthermore, the influence of varying air equivalence ratios, the operation temperature and the fuel moisture on the soot generation in a cyclone gasifier, as well as the ability of the proposed model to reflect such influences, are also discussed. Numerical simulations demonstrate the existence of an optimal operation condition for the cyclone gasifier in terms of the soot generation. The current work thus provides a useful tool for analyzing the mechanism of soot formation at the reactor scale.