Additive manufacturing, particularly 3D printing, has gained significant traction in various industries due to its flexibility and cost-effectiveness. However, one critical aspect of 3D printing is the detection and prevention of defects, as these can compromise the structural integrity and functionality of printed objects. Traditional defect detection methods often rely on manual inspection, which is time-consuming and prone to errors. In recent years, deep learning techniques have shown promising results in automating defect detection tasks, offering faster and more accurate solutions. In this paper, we investigate the effectiveness of various deep learning techniques for the task of 3D printer defect detection. We employ Convolutional Neural Networks (CNNs), Transfer Learning using VGG19, and Recurrent Neural Networks (RNNs) with LSTM and GRU cells. The CNN model achieved the validation accuracy of 90%. By leveraging a pre-trained VGG19 model for feature extraction, this approach effectively captured relevant patterns in the data, demonstrating robust generalization capabilities with an accuracy of 97%. This behavior reflects effective learning and generalization capabilities, indicative of well-adapted model architectures. The RNN model exhibited comparatively lower performance and higher loss values, achieving a validation accuracy of 94% suggesting difficulties in capturing and learning from the sequential patterns inherent in the dataset. The choice of deep learning technique should be carefully considered based on the specific requirements of the application, including dataset size, data complexity, and computational resources. Our experimental results demonstrate the strengths and weaknesses of each technique, providing insights into their applicability to this specific domain including, medical, industrial, or other fields.