The International Journal of Creative Research Thoughts (IJCRT) is an international open-access, peer-reviewed journal. This paper presents a study on cyberbullying detection using deep learning models. Cyberbullying is a serious issue that has become a global epidemic, with harmful messages sent through social media, instant messaging, or digital messages. It can have severe consequences, including emotional and physical harm, and even suicidal attempts. Existing methods for detecting cyberbullying have limitations, such as focusing on a single platform, addressing only one topic, or relying on handcrafted features. Deep learning models can overcome these limitations by learning from one dataset and transferring knowledge to others. The study uses real-world datasets, including Twitter, to evaluate the performance of deep learning models for cyberbullying detection. The paper discusses various deep learning models, including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BLSTM). These models are effective for text classification and can capture important semantic information in text. CNNs are used for image and text classification, while LSTMs are used for learning long-term dependencies in sequence prediction problems. BLSTMs improve model performance by processing input in both forward and backward directions. The study evaluates the performance of these models using confusion matrices, showing that CNNs have a high true positive rate (0.8516) and a low false positive rate (0.3765). LSTMs also show good performance, with a true positive rate of 0.7101 and a low false positive rate of 0.3717. The proposed system uses word embedding and deep learning models to detect cyberbullying in text. The system is efficient and can be applied to different social media platforms. The results show that deep learning models outperform traditional machine learning models in detecting cyberbullying. The study concludes that the proposed system is an effective solution for detecting cyberbullying on social media platforms.The International Journal of Creative Research Thoughts (IJCRT) is an international open-access, peer-reviewed journal. This paper presents a study on cyberbullying detection using deep learning models. Cyberbullying is a serious issue that has become a global epidemic, with harmful messages sent through social media, instant messaging, or digital messages. It can have severe consequences, including emotional and physical harm, and even suicidal attempts. Existing methods for detecting cyberbullying have limitations, such as focusing on a single platform, addressing only one topic, or relying on handcrafted features. Deep learning models can overcome these limitations by learning from one dataset and transferring knowledge to others. The study uses real-world datasets, including Twitter, to evaluate the performance of deep learning models for cyberbullying detection. The paper discusses various deep learning models, including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BLSTM). These models are effective for text classification and can capture important semantic information in text. CNNs are used for image and text classification, while LSTMs are used for learning long-term dependencies in sequence prediction problems. BLSTMs improve model performance by processing input in both forward and backward directions. The study evaluates the performance of these models using confusion matrices, showing that CNNs have a high true positive rate (0.8516) and a low false positive rate (0.3765). LSTMs also show good performance, with a true positive rate of 0.7101 and a low false positive rate of 0.3717. The proposed system uses word embedding and deep learning models to detect cyberbullying in text. The system is efficient and can be applied to different social media platforms. The results show that deep learning models outperform traditional machine learning models in detecting cyberbullying. The study concludes that the proposed system is an effective solution for detecting cyberbullying on social media platforms.