Expected to begin its operations in the coming years, air taxi aims to provide everyday transportation services to customers in metropolitan cities. These vehicles offer urban air mobility (UAM) services using the electric vertical takeoff and landing (eVTOL) technology. This research is the first to present a hybrid simulation goal programming (HSGP) approach to dispatch vehicles in a centralized air taxi network. After each customer drop-off, the model makes real-time decisions on (i) whether the air taxi must become idle or pick up customers, and (ii) the station to which the air taxi should be dispatched (if the air taxi is operational). The feasibility of the HSGP approach is tested using potential air taxi demands in New York City (NYC) provided by a prior study. The results of the experimentation suggest that the minimum number of air taxis required for efficient operation in NYC is 84, functioning with an average utilization rate of 66%. In addition, the impacts of commuter’s “willingness to fly” rate, percentage of demand fulfillment, on-road travel limit, maximum customer wait time, and arrival distribution on the optimal number of air taxis, utilization rate, number of customers served, and cost incurred per customer are examined. Analyses show that the “willingness to fly” rate appears to have a linear influence on the number of air taxis and the efficiency, while on-road travel distance has an exponential impact on the performance measures. The HSGP algorithm developed in this paper can be used by any company that is interested in venturing into the air taxi market.