Lithium-ion (Li-ion) batteries with NMC (Nickel Manganese Cobalt) cathodes require a SOC (state-of-charge) estimation algorithm since it cannot be measured directly. Luenberger-based nonlinear observer is one of the alternative algorithms to estimate the SOC. It corrects the SOC estimation value by multiplying a suitably tuned gain by the difference between the measured terminal voltage and the terminal voltage of the battery model. A correct greater gain would allow the estimation value to reach the steady SOC estimate value faster. However, the higher gain value would magnify the measurement noise, interfering with the calculated SOC value. This work proposes a nonlinear observer-based SOC estimation with added dynamic gain to provide steady SOC estimation quickly while less sensitive to measurement noise. Some numerical simulation results in Python using Li-ion NMC battery parameters show the effectiveness of the proposed observer.