The global shortage of medical isotopes is a common challenge faced by humanity. In-reactor irradiation is the main method of medical isotopes production. Due to the difficulty of precisely regulating the reactor neutron energy spectrum, there is a huge gap between the actual efficiency and the theoretical maximum efficiency of medical isotopes production. We temporarily do not know the theoretical yield limits of various medical isotopes, leading to the impossibility of quantifying this gap and the potential for improving production efficiency. We combined genetic algorithms and burnup algorithms to explore the theoretical yield limits of 20 medical isotopes (14C, 32P, 47Sc, 60Co, 64Cu, 67Cu, 89Sr, 90Y, 99Mo, 125I, 131I, 153Sm, 161Tb, 166Ho, 177Lu, 186Re, 188Re, 92Ir, 225Ac, 252Cf), considering flux levels ranging from 1012 to 1017 (cm-1×s-1-2), and irradiation times from 5 days to 200 days, providing information such as maximum yield, transmutation rate of nuclides, nuclide abundance, and required irradiation time for 8400 (20 × 35 × 12 combinations) scenarios. These data show the limits of nuclides transmutation under the current irradiation conditions, quantify the gap between the current production efficiency and the theoretical maximum production efficiency, and help readers quickly estimate the yield and economy of a reactor for the medical isotopes production.