The accurate detection of colon cancer from histopathology images poses a significant challenge in the field of medical image analysis due to the complexity and diversity of tissue structures. The precision of image-based classification is critical for early diagnosis and effective treatment planning. This research explores the use of advanced deep learning techniques, specifically employing an ensemble of state-of-the-art models, including DenseNet169 and GoogleNet (InceptionV3), to address these challenges. The ensemble method enhances classification performance by leveraging the strengths of multiple models. Using the Colorectal Cancer Histology MNIST dataset, consisting of 5000 histopathology images across 8 classes, this study achieved an impressive accuracy of 88.56%. The integration of DenseNet’s dense connectivity for improved feature extraction and GoogleNet’s inception modules for multiscale processing contributed to the model’s robust performance. Advanced techniques such as early stopping, and data augmentation further optimized the models. The results demonstrate that the ensemble approach offers significant improvements in sensitivity, specificity, and overall classification accuracy for colon cancer detection. This research underscores the potential of deep learning in medical imaging to revolutionize cancer diagnosis, ultimately improving patient care and outcomes.