Google has introduced LiteRT.js, a JavaScript binding for its on-device inference library, formerly known as TensorFlow Lite. This new tool allows .tflite models to run directly within web browsers, leveraging WebGPU, WebNN, and WebAssembly backends. Google claims LiteRT.js offers significant performance improvements, being up to three times faster than other web runtimes for computer vision and audio tasks, and providing 5-60x speedups for demanding real-time applications when utilizing GPU or NPU acceleration. AI
IMPACT Enables more complex AI models to run directly in web browsers, enhancing privacy and reducing latency for web applications.
RANK_REASON This is a software tool release, not a frontier model or significant industry event.
- LiteRT.js
- Apple Silicon
- LiteRT
- ML Drift
- PyTorch
- TensorFlow.js
- TensorFlow Lite
- WebAssembly
- WebGPU
- WebNN
- XNNPACK
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
