Lip reading using artificial intelligence aims to convert visual speech movements into readable text without relying on audio signals. This project presents an AI-based lip reading system that predicts spoken text from video input using deep learning techniques. The system processes uploaded video files and extracts visual features of lip movements, which are analyzed using pre-trained models such as LipNet with LSTM and CTC-based decoding. A Streamlit-based web interface is developed to allow users to upload videos and view the predicted text output easily. The proposed system demonstrates the ability to recognize and translate visual speech into characters, making it useful for assisting hearing-impaired individuals and speech recognition in noisy environments.