Video captcha is a state-of-the-art captcha that modern day a content-based video labelling task to distinguish human beings from bots by providing the captcha to maintain the security, it’s far primarily based on the assumption that people can recognize the content today’s video better than bots, especially when the video is distorted, noisy, or masked. However, it faces the challenge of balancing usability and security. In this paper, proposing a method for implementing video CAPTCHA which integrates balancing and security techniques, including natural conversation, adaptive difficulty, diverse and dynamic video sources. Through user testing and attack simulations, we demonstrate that our approach achieves levels of usability and security compared to existing CAPTCHAs which are not advanced and are text and image based. Additionally, the paper discusses the generation and assessment of video CAPTCHAs using YouTube videos and tags, evaluating system performance and user satisfaction in comparison to traditional CAPTCHA methods.