This study presents a novel approach to stock price prediction by enhancing the Extreme Learning Machine (ELM) with multi-indicator fusion. Our method extracts nine technical indicators from historical stock price data, combining them to significantly enhance forecasting accuracy. The dataset includes stock data from NYSE and NASDAQ spanning several years, covering eight firms: GS, MSFT, AAPL, IBM, CTSH, BAC, ORCL, and HAL. Our approach is rigorously evaluated through a comprehensive comparative study using six key performance metrics: Average Relative Value (ARV), Mean Absolute Error (MAE), Symmetric Mean Absolute Percentage Error (SMAPE), Root Mean Squared Error (RMSE), Mean Squared Error (MSE), and \({{{R}}}^{2}\) , applied across multiple datasets. We compare the effectiveness of our ELM-based approach with several well-known machine learning models, including Artificial Neural Networks (ANN), Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), and Multi-Layer Perceptron with Backpropagation (MLP-BP). The results show that the ELM model, enhanced with a new combination of technical indicators, outperforms competing methods in prediction accuracy and robustness. This study highlights the efficacy of ELM in stock price forecasting and contributes to the growing research on integrating diverse data sources with machine learning techniques for financial predictions.