The volatility of an individual stock plays an important role in the pricing and risk management of various financial instruments. Volatility does not behave like stock price, which is close to a random walk, on the contrary, volatility has its own mathematical and statistical properties. In this paper, I modeled and forecasted the volatility of selected US stocks by using Generalized Autoregressive Heteroscedasticity (GRACH) and the variations of the GARCH model. I showed the fitting result, prediction, and loss function of each model and concluded their strengths and weaknesses.