Core ML

Integrate machine learning models into Apple apps with a high-performance framework optimized for on-device processing.

Core ML serves as the foundational framework for deploying machine learning models across iOS, macOS, watchOS, and tvOS. It leverages the Apple Neural Engine, GPU, and CPU to minimize memory footprint and power consumption while maintaining user privacy through local execution. Developers can convert models from popular libraries like PyTorch and TensorFlow using the coremltools converter or build custom solutions with Create ML. The framework supports advanced features (including on-device model personalization and asynchronous prediction) to ensure smooth performance in apps like Photos, Siri, and Camera.

https://developer.apple.com/documentation/coreml