LiteRT

LiteRT, successor to TensorFlow Lite. is Google's On-device framework for high-performance ML & GenAI deployment on edge platforms, via efficient conversion, runtime, and optimization

21 articles 1.8k View on GitHub ↗
21 articles
LiteRT Security Features: Defense-in-Depth for On-Device Machine Learning

Explore LiteRT security features including defense-in-depth, sandboxing for untrusted models, and strict input validation for on-device ML. Protect your applications effectively.

deep-dive
Mar 13, 2026
How to Optimize Model Size with LiteRT: Quantization and External Weights Explained

Optimize model size with LiteRT using quantization and external weights. Reduce footprints up to 4x or separate large weights from metadata.

how-to-guide
Mar 13, 2026
LiteRT Ecosystem Tools: Complete Guide to Command-Line Utilities for Model Deployment

Explore the LiteRT ecosystem and its 13 command-line tools for deploying, analyzing, and debugging TensorFlow Lite models on CPU, GPU, and NPU hardware. Get started with google-ai-edge/LiteRT today.

deep-dive
Mar 13, 2026
LiteRT Usage Examples: Complete Guide to On-Device Inference

Explore LiteRT usage examples in the official repository including interpreter demos and CompiledModel API implementations for on-device inference on hardware accelerators.

tutorial
Mar 13, 2026
How to Contribute to the LiteRT Project: A Complete Developer Guide

Learn how to contribute to the LiteRT project by following this developer guide. Sign the CLA, fork the repo, build locally, and submit your PR for our AI edge computing project.

how-to-guide
Mar 13, 2026
LiteRT Known Issues and Limitations: Critical Constraints for Edge AI Deployment

Discover LiteRT known issues and limitations including stateless LSTM support, GPU operator whitelists, and NPU API version matching. Optimize your edge AI deployment.

deep-dive
Mar 13, 2026
How to Debug Applications Using LiteRT: A Complete Guide

Learn how to debug applications with LiteRT. This guide details using command-line tools and the LiteRtProfilerT API to inspect tensors, find errors, and profile performance across CPU, GPU, and NPU.

how-to-guide
Mar 13, 2026
Does LiteRT Support On-Device Training? Current Capabilities and Future Roadmap

LiteRT currently offers inference-only runtime. Discover its capabilities and explore the planned roadmap for future on-device training support in this technical overview.

deep-dive
Mar 13, 2026
LiteRT Performance Benchmarks: Measuring Latency, Memory, and Throughput

Discover LiteRT performance benchmarks. Measure latency memory footprint and throughput of TensorFlow Lite models with LiteRTs C++ benchmark tool. Optimize your edge AI performance now.

performance
Mar 13, 2026
How to Compile LiteRT from Source: Docker, CMake, and Bazel Guide

Compile LiteRT from source using Docker, CMake, or Bazel. Follow our guide for reproducible containerized builds and cross-compilation for Android. Get started easily.

how-to-guide
Mar 13, 2026
Hardware Platforms Supported by LiteRT: CPU, GPU, and NPU Coverage

Explore LiteRT hardware support including CPU GPU and NPU acceleration across Android iOS Linux macOS Windows Web and IoT devices optimizing on-device ML inference.

api-reference
Mar 13, 2026
Can LiteRT Be Used for Real-Time AI Applications? Architecture and Implementation Guide

Explore LiteRT for real-time AI applications. Discover how its architecture delivers sub-30ms inference on edge devices with asynchronous acceleration and efficient memory management.

architecture
Mar 13, 2026

Have a question about this repo?

These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:

Share the following with your agent to get started:
curl -s "https://instagit.com/install.md"
Works with
Claude Codex Cursor VS Code OpenClaw Any MCP Client

Maintain an open-source project? Get it listed too →