Wind energy is a crucial component of the ongoing energy transition. Over the past decade, technologies in this field have advanced significantly. Currently, the industry is following a trend of increasing the size of wind turbines, both onshore and offshore units. However, this trend brings about complexities in design, maintenance, and operation. Wind turbines are highly dynamic structures, and accurate prediction of dynamic loads is crucial for future design and, consequently, installation costs and lifetime expectancy. Dynamic stall is a highly dependent process influenced by various parameters such as Reynolds number and airfoil geometry. It involves significant changes in pressure in the flow field around the airfoil due to the development and shedding of vortex structures from the leading or trailing edge. Another characteristic is flow reversal from the trailing edge, leading to a delay in boundary layer attachment and the formation of significant hysteresis in lift force, pitching moment, and drag force values. Due to the constant rotation of the rotor, this process is cyclic, necessitating accurate modeling for proper assessment of fatigue damage, instability levels, and turbine noise. This study aims to compare the results of dynamic stall models for the FFA-W3-211 airfoil, with a maximum thickness of 21.1% of the chord length. Previously unavailable dynamic data for the considered airfoil is obtained through experiments in a Low-Speed Low-Turbulence Wind Tunnel located at TU Delft. Static and dynamic polars are obtained for various Reynolds numbers, amplitudes, mean angles, and oscillation frequencies. Measurements are conducted using pressure taps located on the airfoil model, and static data is validated against existing data and RFOIL results. The study considers two semi-empirical models. The first is the Beddoes-Leishman model, one of the most popular at the time of writing. Its optimization and comparison with the default version revealed that optimized shape-based coefficients are crucial for obtaining correct results. It is found that the optimized model significantly outperforms the default version in almost all cases. Additionally, it is established that the model lacks the leading edge vortex effect in modeling negative stall. The second model chosen is the first-order Snel model. Optimization of this model also led to significant improvements in the results obtained. A significant improvement in modeling both hysteresis size and reattachment area with increasing Reynolds number is identified. However, a significant drawback of the model in modeling negative stall is revealed. During uppstroke motion, the model predicts trailing edge vortex effects instead of leading edge, which negatively affects the overall model performance. Comparison of the two models identified their strengths and weaknesses. The Beddoes-Leishman model appears more stable for all cases considered, while the Snel model is sensitive to changes in Reynolds number. It is also noted that both models have difficulties in correctly determining dynamic stall onset angle of attack and the slope of the normal force coefficient. Additionally, both models tend to overpredict values in a greater number of cases. When comparing negative stall, the optimized Beddoes-Leishman model performs significantly better than the Snel model. It accurately predicts the overall form and severity of hysteresis as well as the reattachment area. The Snel model requires additional modifications for correct modeling of negative stall.