The rise of deep learning models has transformed numerous domains, including personalized search. This approach customizes search results based on user preferences, behavior, and context. Leveraging neural networks, deep learning models can analyze vast datasets, uncover complex patterns, and enhance search relevance. In this study, we explore key deep learning methods for personalized search, evaluating their performance across two datasets in various conditions to achieve optimal results. Our assessment includes convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformers, offering a detailed comparison of their strengths and limitations. By conducting a comprehensive evaluation, we aim to refine the state-of-the-art in personalized search, identifying the most effective models and addressing practical implementation considerations. Our findings provide valuable insights for researchers and practitioners, guiding future advancements in deep learning-powered personalized search and improving the delivery of highly relevant results to users.