Aspect-based sentiment analysis is a task aimed at understanding the sentiment expressed in opinion texts or reviews about specific features of an entity, and is currently playing a key role in a variety of scenarios. The automatic extraction of aspects (characteristics) is the most challenging task, as it requires the ability to understand the context and to recognize the relevant and characteristic elements of an entity about which you have an opinion. To increase the quality of the solution to this problem is still a challenge in Spanish, because very few papers have been reported and the reported efficacy rates need to be improved. In this work, an aspect extraction method in which several language Pre-trained Transformer Models are combined through an ensemble approach is presented. The proposed solution was evaluated using the SemEval2016 review dataset and the results obtained were compared to those reported by other state-of-the-art solutions. The evaluation process developed not only provides a starting point to have a broader perception of the performance of the Transformers models in the solution of this problem, but also highlights the improvement of the quality of the results with the combination of models through ensemble techniques.