Phishing is a major cybersecurity threat affecting individuals and organizations worldwide. Attackers use malicious URLs to trick users and steal sensitive information like login credentials, financial details, etc. Traditional detecting systems such as blacklist and heuristic-based methods were struggling to keep up with the rapid evolution of phishing techniques. This study presents a deep learning-based approach for phishing URL detection, with multiple deep learning architectures which includes an Artificial neural network (ANN). Conventional neural network (CNN), recurrent neural network (RNN), and hybrid models like CNN+ANN, CNN+RNN, and CNN+ANN+RNN. This model were trained and evaluated using a dataset, which consisting of legitimate and phishing URL features, which are represented as 0 and 1 for the classification. The model's performances were assessed using key metrics such as accuracy, precision, recall, F1-score. The results demonstrated that the hybrid model gives good accuracy, compared to individual deep learning models, achieving higher accuracy and robustness in detecting phishing attempts. The study highlights that the effectiveness of deep learning techniques in detecting phishing threats and continuous model improvement to prevent emerging attack strategies.