Electricity market decarbonization using renewable energies is essential to significantly reduce greenhouse gas emissions. However, the identification of the optimum electrical mix for each country is a challenging and complex task because of the many boundary conditions that must be taken into account, thus demanding the use of advanced analysis and computing tools. This paper presents a methodology based on artificial intelligence developed for this purpose and the results obtained when applied to the specific case of the Spanish electricity market long-term decarbonization. It also shows how effective this methodology is to find the optimum electrical mix fulfilling previously defined objectives: 100% coverage of the demand while simultaneously minimizing the electricity cost and curtailments. Results show the significant barriers to achieve a 100% renewable electrical mix without excessive curtailments or installed power. The methodology described here is suitable for any other country and target objectives. •Smart electricity mix optimization based on artificial intelligence.•Multi-objective optimization to simultaneously minimize cost and curtailments.•The optimized mix covers the demand without exceeding the maximum greenhouse gas emissions.•Application to the long-term full decarbonization of the Spanish power generation.