This paper proposes method for solving steel alloying problem using evolution algorithms SOMA and differential evolution. Both algorithms belong to the family of the evolution algorithms but the main ideas of these algorithms are different. In differential evolution new offspring is created during the evolution, the individuals are crossed and mutated, while in SOMA the individuals move in the space of the possible solutions and the mutation is replaced by perturbation. The main goal of this paper is to discover how much these algorithms are usable and suitable to solve the problem of steel alloying.