# API reference

heliaRT uses the LiteRT for Microcontrollers model, interpreter and resolver API. Begin with the [runtime API map](https://ambiqai.github.io/helia-rt/reference/runtime/) to find each declaration and its header.

For the inference sequence and error handling, follow [First inference](https://ambiqai.github.io/helia-rt/getting-started/first-inference/). For source-versus-prebuilt integration, [choose your path](https://ambiqai.github.io/helia-rt/getting-started/choose-your-path/).

## Upstream foundation

heliaRT is built on [LiteRT for Microcontrollers](https://github.com/tensorflow/tflite-micro). Upstream-derived files retain their attribution and licenses. Consult [the repository license](https://github.com/AmbiqAI/helia-rt/blob/main/LICENSE) and each component's notices when distributing software.