Stocks are a vital component in the global financial system because they not only provide alternative financing for companies, but also become investment instruments for individuals and institutions. stock prices tend to be volatile and are influenced by many external factors, so forecasting methods are needed that can effectively handle these dynamics and uncertainties. Conventional approaches such as technical and fundamental analysis still have limitations in overcoming complex and non-linear data patterns. Therefore, this research uses Cheng's Fuzzy Time Series method as a more flexible alternative in analyzing and predicting stock prices. This research focuses on Apple Inc. stock, which is one of the issuers with the largest market capitalization in the world. Historical data is taken from the Kaggle platform with a time span of 1980 to 2025, but the period used for forecasting starts from January 2020 to June 2, 2025. The prediction process was carried out for June to December 2025, using three configurations of the number of intervals, namely 7, 9, and 11, to determine the effect of interval selection on the accuracy of the model. Based on the evaluation results, the use of 11 intervals provides the best performance with a MAPE value of 2.36% and an RMSE of 6.28. Furthermore, long-term forecasting until December 2026 was also conducted using the same intervals, and the results showed that the model maintained a stable level of accuracy.