Anemia has become one of the global public health burdens that distracts the well-being of people and affects billions of people across the world. Anemia occurs when the hemoglobin level within the body is reduced below the normal threshold, or when there is a deficiency in the number of blood cells the body produces. The clinical method for anemia diagnosis is centered on the invasive mechanism, that is, extraction of blood, which is a time-consuming technique and can also expose medical personnel to blood-transmissible diseases or infection and results in greater labor and cost of equipment. A hybrid three-layer convolutional neural network architecture is utilized to detect anemia using conjunctiva images. Three single convolutional neural network architectures, that is EfficientNetV2B0, MobileNetV3Small and ResNet50V2 are integrated to develop a three-layer hybrid CNN architecture (EfficientNetV2B0 + MobileNetV3Small + ResNet50V2) to detect anemia. The model exhibits superior detection performance with enhanced computational efficiency. The results of this study highlight the trade-offs between accuracy, robustness, and computational efficiency in deep learning models for anemia detection, which also provides a significant accuracy, for real-world applicability in clinical settings and deployment. This study emphasizes the transformative role of deep learning in advancing anemia detection, paving the way for scalable, accurate, and computationally efficient tools in modern healthcare applications.