Intelligent farming has emerged as a revolutionary approach to modern agriculture, particularly in greenhouse environments, where controlled conditions enhance productivity and ensure sustainability. Advanced technologies enable greater automation in greenhouse management, optimizing resource utilization and minimizing human intervention. Model Predictive Control (MPC) is a key control strategy in greenhouse systems, coordinating various components such as sensors, actuators, and energy management systems. However, with the advent of Artificial Intelligence (AI), integrating AI-based predictive models into MPC can enhance the efficiency and intelligence of greenhouse management. This paper presents a comparative analysis of the Long Short-Term Memory (LSTM) model for time-series forecasting using different input sequence lengths. The objective is to assess the performance of LSTM with various window sizes to identify the most suitable configuration for deployment in an MPC system for 24-h ahead climate forecasting. Accurate forecasting of external weather conditions enables proactive adjustments to the internal greenhouse environment, reducing energy consumption while maintaining optimal growing conditions. By anticipating climate variations, the system can dynamically optimize climate variables within the greenhouse, such as temperature, humidity, and CO₂ concentration, thereby minimizing energy waste and improving overall efficiency. The performance of the forecasting models was evaluated using multiple metrics, including Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Symmetric Mean Absolute Percentage Error (SMAPE). This study highlights the potential of AI-driven predictive control in greenhouse automation, contributing to more sustainable and energy-efficient agricultural practices.