Hyperspectral image (HSI) classification using convolutional neural networks often grapples with a large number of network parameters and a scarcity of class-labeled samples. To tackle these issues, we propose a method called AL-CGNet, which integrates active learning and clustering with group convolutions network for efficient HSI classification. AL-CGNet combines a convolutional neural network with active learning and clustering to enhance feature extraction and classification, while a group convolutions-based lightweight network model significantly reduces parameter count. Initially, HSI reduced in dimensionality through linear discriminant analysis is segmented into clusters via the mini-batch K-means algorithm. The central feature of each cluster substitutes the samples within, leveraging information from unlabeled samples. Subsequently, feature maps are segmented into groups along the spectral dimension in the group convolutions network, where each group sequentially extracts spatial-spectral features through multiple residual blocks. This grouping strategy optimizes band redundancy and diversity, cuts down network parameters, and achieves lightweighting. Active learning then selects informative samples for the training set, mitigating the issue of limited labeled samples. Experimental results demonstrate that AL-CGNet, with only 6% training samples, significantly outperforms ClusterCNN, SSRN, and HybridSN on the Indian Pines, Botswana, and Houston datasets, achieving overall accuracies of 99.57%, 99.23%, and 98.82%, respectively. Remarkably, AL-CGNet remains effective even with a smaller training sample size of 5%. This method not only boosts HSI classification efficiency but also ensures robust feature extraction and high accuracy.