Colorectal cancer is one of the leading causes of cancer-related mortality worldwide, necessitating efficient and accurate diagnostic approaches. This study proposes a deep learning-based ensemble model for the detection and classification of colorectal cancer using histopathological images from the CRC-VAL-HE-7K dataset, comprising 7,190 images across nine classes. Four state-of-the-art convolutional neural networks (CNNs) were employed: ResNet-50, VGG-19, MobileNet V4, and Inception V4 for feature extraction and classification, with predictions aggregated using a majority voting strategy. The ensemble model achieved superior performance, with an accuracy of 94.31%, a precision of 95.30%, a recall of 90.12%, and an F1 score of 92.54%, outperforming individual models such as ResNet-50 and Inception V4, which achieved accuracies of 89.45% and 91.25%, respectively. These results highlight the ensemble model’s effectiveness in improving diagnostic generalization and robustness. Furthermore, the proposed approach has significant clinical implications, offering potential integration into computer-aided diagnostic systems to support faster and more accurate detection of colorectal cancer. By reducing reliance on subjective interpretation by pathologists, this model can facilitate early detection, streamline diagnostic workflows, and ultimately improve patient outcomes.