LiteRT is Google’s next-generation runtime for running AI models efficiently on edge devices. Built as the successor to TensorFlow Lite, it enables faster, optimized, and on-device AI experiences across mobile, embedded, and IoT systems.