Rigorous assessment and testing approaches are vital for promoting effective learning. However, traditional methods often fall short of capturing the dynamic and nuanced nature of users’ abilities. To overcome these limitations, adaptive testing systems are being employed. They have revolutionized assessment in educational and psychological fields through personalized learning experiences that are tailored to an individual’s needs and abilities. Central to this system’s success are the question difficulty and user ability parameters, which govern the selection and sequencing of testing materials. In this paper, we propose innovative mathematical models of question difficulty and user ability estimation for adaptive testing systems. Leveraging user response time, correctness of responses, adjustable weights, and relative performance, we present a novel approach for the precise gauging of question difficulty and user ability parameters. We also demonstrate the efficacy of our models through empirical evaluations and highlight their advantages and disadvantages concerning the existing models.