Google Launches LiteRT Tensor API for Building, Customizing, and Optimizing ML Models for On-Device

Google launched the LiteRT Tensor API, a lightweight C++ interface enabling developers to build and optimize ML models for on-device execution across CPU, GPU, and NPU without rewriting code for each hardware target. It supports direct weight loading from checkpoints and zero-copy buffer interoperability, demonstrated with a multi-object segmentation app running faster than Apple MLX on MacBook Pro.