In this work, optimal design for solar steam methane reforming using molten salt (SSMRMS) with CO2 capture and utilisation is investigated by employing a machine-learning based optimisation framework. The results obtained show that a 65.77 % reduction can be obtained compared with the existing SSMR-MS. The optimal Levelised Cost of Hydrogen Production (LCHP) is 1.26 $ kg -1 which represents about half of the LCHP in the existing SSMR-MS. The captured CO2 can produce around 485.73 kt polypropylene carbonate annually.