In recent years, there has been considerable interest in the development of zero-emissions, sustainable energy systems utilising the potential of hydrogen energy technologies. However, the improper long-term economic assessment of costs and consequences of such hydrogen-based renewable energy systems has hindered the transition to the so-called hydrogen economy in many cases. One of the main reasons for this is the inefficiency of the optimization techniques employed to estimate the whole-life costs of such systems. Owing to the highly nonlinear and non-convex nature of the life-cycle cost optimization problems of sustainable energy systems using hydrogen as an energy carrier, meta-heuristic optimization techniques must be utilised to solve them. To this end, using a specifically developed artificial intelligence-based micro-grid capacity planning method, this paper examines the performances of twenty meta-heuristics in solving the optimal design problems of three conceptualised hydrogen-based micro-grids, as test-case systems. Accordingly, the obtained numeric simulation results using MATLAB indicate that some of the newly introduced meta-heuristics can play a key role in facilitating the successful, cost-effective development and implementation of hydrogen supply chain models. Notably, the moth-flame optimization algorithm is found capable of reducing the life-cycle costs of micro-grids by up to 6.5% as compared to the dragonfly algorithm. Display omitted •A metaheuristic-driven model is proposed to optimally design H2-based microgrids.•An energy management strategy is used to facilitate the uptake of fuel cell vehicles.•The performances of twenty metaheuristics are studied in terms of solution quality.•The moth-flame optimizer outperforms the well-respected algorithms in this area.•Levelized costs of electricity and H2 are found to be well below the present rates.