Temporal knowledge graph embedding (TKGE) models play a crucial role in representing entities, relationships, and timestamps, enabling the inference of missing facts in dynamic environments. However, current models face a significant challenge in distinguishing representations of the same entity at different time points, limiting their ability to predict links accurately in temporal knowledge graphs (TKGs). In this paper, we introduce Time-Aware Complex Attention Space (TACAS), a novel and straightforward model that directly incorporates temporal information into entity embeddings, allowing entities to evolve over time. This evolution enables TACAS to capture the temporal dynamics of relationships with greater precision. Moreover, TACAS leverages an attention mechanism within a complex space framework, which enhances the interaction between entities, relations, and time, improving the model’s sensitivity to temporal information. Our experiments on four benchmark datasets demonstrate that TACAS outperforms majority of existing TKGE models.