In the real-world systems, the interactions between objects (e.g., molecules) are generally represented through the dynamic networks, based on a graph-model that evolve over the time. Network alignment allows evaluating different networks in terms of homology and topology. It is a method to map nodes from different networks, in order to match the same entities. For instance, we can assume that the topological similarity between the regions of two given networks corresponds to their functionality, e.g., in terms of biological process. Therefore, the alignment between the biological networks of two species may be useful to transfer the knowledge from the simplest (e.g., mouse) to the more complex (e.g., human).In this paper, we present DANTE (DynAmic Networks alignment based on Temporal Embeddings), a solution for the alignment of dynamic networks, based on temporal embeddings. DANTE takes into account the similarity between the nodes based on their evolution and relationships between one time point and the subsequent, for all ones in the networks. Our temporal embeddings concern a set of vectors representing the topological similarity between the nodes of two given networks, by representing each node as a word on the Skip-Gram model. In addition, we extend the embeddings process to the dynamic networks, in order to evaluate the similarity between two dynamic networks. DANTE applies an iterative process for maximizing globally the match score between the pair of nodes.DANTE is freely available on https://github.com/pietrocinaglia/dante.