Google Tensor SDK Beta with LiteRT

Source: Google Developers Blog

Overview

The Google Tensor ML SDK has entered its Beta phase, enabling developers to build and deploy high‑performance machine learning models directly onto the TPU of Google Pixel 10 devices.

Key Features

• LiteRT integration – Google’s edge deployment framework provides a unified workflow to convert, compile, and run PyTorch or TFLite models with robust fallback options.
• Unified workflow – Simplifies the process of moving models from development to on‑device execution.

Model Garden

A new model garden now offers over 100 classic and generative AI models, including Gemma 3. These models support low‑latency, private on‑device capabilities such as:

• Speech recognition
• Computer vision
• Text generation