In this article, we have evaluated the dynamics of deflation in Iran^,s economy. We have considered the assumption of boundd rationality,taking into account agent^,s learning and finite horizon planning. The estimation of the models was conducted using the Bayesian approach within the framework of the dynamic stochastic general equilibrium. We used seasonal data from the Iranian economy covering the period 1988:2-2023:1. Our results indicate deflation in the finite horizon model with agent lr=earning exhibit less fluctuations and is more persistent compared to the Canonical New Keynesian model. This is due to the slow learning of agents from past experiences.As a result, deflation requires less aggressively reaction of central bank.