Custom-designed SOCs developed by Google.
Unlock peak performance with Google Tensor that powers Google's Pixel devices.
The Google Tensor SDK is a specialized software toolkit that allows developers to optimize and run machine learning models directly on the Tensor Processing Unit (TPU) found in Pixel devices.
Sign-up for Beta
Be among the first to access the Google Tensor SDK and unlock the full potential of your AI applications on Google hardware.
Run AI models on Google Tensor
Streamline your deep learning workflow from training to on-device deployment.
1. Compilation
Use LiteRT AOT compiler to compile your models for supported SOCs.
2. Deployment
Deploy your models on with Google Play on Android devices.
3. Inference
Run your model on Google Tensor.
120+ models, Ready for Pixel
Bring cutting-edge vision, text, and audio capabilities directly to your Pixel devices. Explore our Model Garden library and use the Google Tensor SDK for seamless, on-device deployment.
Blogs and Announcements
Stay up to date with the latest announcements, technical deep dives, and performance benchmarks from the Google Tensor team.
LiteRT: The universal framework for on-device AI
Google's unified on-device ML framework, evolving from TFLite for high-performance deployment.
Join the Community
Google Tensor GitHub Community
Contribute directly to the project and collaborate with core developers.
Hugging Face Hub
Access optimized open-weight models on the Hugging Face Hub.