OpenAI has entered the hardware market with a device aimed squarely at software developers.
LiteRT.js runs machine learning models locally with CPU, GPU and emerging NPU acceleration, potentially reducing server infrastructure, inference charges and data movement.