Incorporating semantic analysis into image compression can significantly reduce the repetitive computation of fundamental semantic analysis in client-side applications such as semantic image retrieval. The same practice also enables the compressed code to carry semantic information of the image during its storage and transmission. In this paper, we propose a Deep Semantic Image Compression (DeepSIC) model to achieve this goal and put forward two novel architectures that aim to reconstruct the compressed image and generate corresponding semantic representations at the same time by a single end-to-end optimized network. The first architecture performs semantic analysis in the encoding process by reserving a portion of the bits from the compressed code to store the semantic representations. The second performs semantic analysis in the decoding step with the feature maps that are embedded in the compressed code. In both architectures, the feature maps are shared by the compression and the semantic analytics modules. Experiments over benchmarking datasets show promising performance of the proposed compression model.