The increasing prevalence of hate speech, on platforms poses a threat to inclusive communication. In our research we have utilized word embeddings to acknowledge the challenge of identifying Hate speech. For enhancing the accuracy and contextual understanding of our hate speech detection models we have incorporated both BERT (Bidirectional Encoder Representations, from Transformers) and GloVe (Global Vectors for Word Representation) embeddings. By preprocessing input and training machine learning models using these embeddings we have developed a methodology that effectively recognizes hate speech. Our experiments demonstrate the effectiveness of combining GloVe and BERT in detecting instances of hate speech. This proposed approach not outperforms techniques but also captures the intricate context in which hate speech manifests. Through this study we contribute to endeavors aimed at mitigating the impact of Hate speech on platforms.