In this paper we propose a new architecture based on Neural Cellular Automata and evaluate it on the task of binarization of scanned historical documents. We show that this approach allows us to obtain neural models that outperform existing Neural Network-based solutions while exhibiting substantially lower complexity measured as the number of parameters of underlying neural network. The proposed model will be evaluated over several settings including different forms of neighborhood, numbers of steps and initialization strategies. On the basis of this evaluation we can select the most suitable setup in order to produce models exhibiting high performance and quality of results.