The field of time series analysis is witnessing a significant shift towards leveraging large language models (LLMs) to improve performance and efficiency. Recent research has focused on integrating time series data with LLMs, enabling them to handle complex temporal patterns and relationships. This integration has led to the development of novel architectures and frameworks that can effectively reason over multivariate time series data, generating insights and detecting anomalies. Notably, the use of LLMs has shown promising results in handling variable-length time series sequences and context-based anomalies, addressing long-standing challenges in the field. Furthermore, the application of LLMs to time series analysis has also led to the creation of modular frameworks for robust time series decomposition, allowing for more flexible and interpretable analysis. Overall, the incorporation of LLMs in time series analysis is opening up new avenues for research and development, with potential applications in diverse domains such as finance, healthcare, and scientific discovery. Noteworthy papers in this area include OpenTSLM, which introduces a family of Time Series Language Models for reasoning over multivariate medical text- and time-series data, and TS-Reasoner, which aligns the latent representations of time series foundation models with the textual inputs of LLMs for downstream understanding and reasoning tasks. Additionally, SciTS presents a benchmark for scientific time series understanding and generation, highlighting the potential of LLMs in this area. SPEAR and THEMIS also demonstrate the effectiveness of LLMs in anomaly detection, with SPEAR leveraging soft prompts and quantization, and THEMIS exploiting pretrained knowledge from foundation models.