Effective grading of arecanut is essential for ensuring product quality, maximizing market competitiveness, and satisfying consumer preferences. However, traditional methods of arecanut grading are challenging due to variations in arecanut size, shape, and appearance, resulting in subjective and inconsistent evaluations. Deep learning can enhance this process by automating grading and using sophisticated algorithms to assess both visual and non-visual attributes, thereby increasing efficiency, accuracy, and consistency. This study presents two standalone CNN-based methodologies for automated arecanut quality grading, leveraging DenseNet121 and InceptionV3 with custom layers tailored for arecanut classification. A dataset of 2,000 high-resolution images, manually curated from farms and augmented for diversity, was used for training and validation. Eight CNN architectures - DenseNet121, EfficientNetB4, InceptionResNetV2, InceptionV3, MobileNetV2, ResNet50, VGG16, and VGG19 were evaluated. Experimental findings showed DenseNet121 and InceptionV3 achieved the highest accuracy (95.67%) and strong precision/recall scores (96%), making them the most promising models. Meanwhile, MobileNetV2 was identified as the fastest model in terms of classification speed; however, its relatively low accuracy limits its practical application in grading tasks. DenseNet121 and InceptionV3, while marginally slower at 0.015 and 0.011 seconds per image, respectively, offered a good balance between computational cost and elevated accuracy. DenseNet121 excels in feature reuse through its dense connectivity, reducing redundancy and improving performance on smaller datasets, while InceptionV3 utilizes multi-scale feature extraction to capture intricate patterns effectively. Both models demonstrate robustness under varying conditions, ensuring reliability in practical deployment scenarios. This study highlights the potential of CNNs to provide a reliable, and scalable solution for arecanut grading, benefiting farmers by expanding market opportunities.