Large Language Models in Wearable Devices: Applications, Challenges, and Future Directions

The intersection of wearable technology and Large Language Models (LLMs) is enabling new possibilities in health monitoring, fitness coaching, gesture control, and personalized care. This survey presents a comprehensive overview of current research efforts that integrate LLMs with wearable systems to analyze wearable data—including motion sensors, biosignals, and vital signs—to generate personalized insights within wearable platforms such as smartwatches, neural interfaces, and fitness ecosystems. We examine state-of-the-art methods that reflect these trends, focusing on system design, data modalities, and human–AI interaction mechanisms. Key use cases, model architectures, sensor integrations, and application scenarios are reviewed. We also discuss challenges related to privacy, real-time processing, personalization, and the feasibility of on-device inference.