This study suggests a neural network (NN) algorithm to accurately guess the state of charge (SOC) of lithium-ion batteries. This is important for fixing range anxiety in electric cars and making portable electronics work better. By leveraging real-time data, such as current, voltage, and temperature, the proposed model accounts for the dynamic variations occurring during charge and discharge cycles. The model is developed using experimental data collected in a controlled laboratory environment. The Levenberg-Marquardt algorithm (LM) is utilized in Matlab to train the NN algorithm and optimize its parameters under various operational scenarios. Rigorous performance evaluations are conducted, comparing the model's predictions with actual measurements. With 30 hidden neurons, the best-performing model achieved an impressive performance ratio of 19.66. These results demonstrate the reliability and accuracy of the proposed NN approach in predicting battery SOC. This method offers promising potential for improving battery management across various applications, including electric vehicles, renewable energy storage systems, and portable electronics. By doing so, it contributes to advancing sustainable and innovative energy solutions.