International trade serves as a significant pillar of economic growth, particularly through export activities that contribute to national income. Among Indonesia’s creative products, batik holds both economic and cultural value, making it a strategic commodity in the global market. This research aims to forecast the export value of Indonesian batik to the United States, Japan, and Germany until 2045. To achieve this, two forecasting approaches were utilized: Autoregressive Integrated Moving Average (ARIMA) and Long Short-Term Memory (LSTM). The study employs annual export data spanning from 2010 to 2021. The comparison of both models reveals that ARIMA provides more reliable forecasts, supported by lower Mean Absolute Percentage Error (MAPE) scores. In contrast, the LSTM model encountered difficulties in handling the dynamic patterns of the data, likely due to the limited time span and fluctuations in export values. The findings highlight ARIMA’s strength in capturing long-term trends, making it a more suitable method for this case. These insights are expected to assist policymakers and stakeholders in designing effective strategies to support batik exports as part of Indonesia’s long-term development vision. Future research is suggested to include broader datasets and integrate external economic variables to enhance the performance of deep learning-based forecasting models.