Free-floating bike-sharing is characterized by environmental-friendly and a sustainable form of urban transportation, which becomes increasingly popular globally. To achieve sustainable development, we establish an analytical framework to explore the optimal business models of free-floating bikesharing platforms (FBSPs) including ordinary, premium, and hybrid that are three business models often employed by FBSPs. This paper aims to address the platform’s optimal business decision, and to characterize the conditions under which business model can achieve Pareto improvement for the platform and users. Our results reveal that both user’s heterogeneity on service quality sensitivity and time for searching a bike play critical roles in deciding platform’s optimal business model. If the platform only provides premium service, high operating costs lead to lower profit than that of hybrid strategy. Furthermore, the hybrid service model is optimal when platform overall performance (the ratio of the time to searching a bike and the platform service quality) is either sufficiently high or low, and Pareto improvement for the platform and users are also achieved under this condition. Otherwise, the ordinary service model should be implemented when platform overall performance is moderate. Service quality has also found having a non-monotonic effect on social welfare under premium service model.