Abstract In recent years, significant research efforts have focused on increasing power extraction using wake steering through yaw misalignment. In the current work, we evaluate the potential increase in annual energy production (AEP) for the Belgian offshore wind farm cluster using the FLORIS framework. The 2.2 GW Belgian cluster is an excellent test case to explore cooperation between wind farms as the concession zone is split into nine individual farms, resulting in a mix of intra- and inter-farm wakes to be considered in one of the largest wind farm clusters operational to date. Furthermore, the cluster high power density with relatively low turbine spacings is found to be responsive to wake mitigation strategies. Three different cost functions are compared. The first one is the complete cluster yaw angles optimization. The other two are farm-based optimizations seeking to maximize either their own AEP or the cluster AEP. Results show that all wind farms benefit from every optimization, encouraging operators to implement this strategy. The cluster optimization is found to be the one providing the best AEP increase for all wind farms individually and thus for the cluster. The farm-based function optimizing the cluster AEP is in second place, encouraging wind farm operators to collaborate for their own profits.