Text summarization is a crucial aspect of natural language processing, finding widespread applications in various aspects of our lives. Numerous studies have delved into this task, employing diverse approaches ranging from linguistic analysis to machine learning and deep learning methodologies. However, there has been limited exploration of transformer-based models for Vietnamese text summarization. In our research, we leverage transformer models, namely, PhoBERT, BARTpho, and ViT5, to develop and refine effective summarization models. Through experimentation on the Vietnews dataset and evaluation using ROUGE metrics, our proposed method demonstrates noteworthy achievements in the realm of Vietnamese text summarization, showcasing its potential for advancing the quality of summarization tasks.