A new open-source tool, coreml-cli, provides a native command-line interface for working with Apple Core ML models on macOS. It allows developers to inspect, run inference, benchmark, and manage models without needing Xcode or Python. The tool supports batch processing, compiling models to the `.mlmodelc` format, and selecting compute devices like the Apple Neural Engine.

This addresses a gap for developers who prefer scripting and automation over GUI-heavy workflows. The CLI streamlines common tasks like performance benchmarking and CI/CD integration, offering JSON output for easy parsing. It’s built with Swift and requires macOS 13+, aligning with modern Apple development environments.

The project is available via Homebrew or direct download from GitHub. Its release suggests a growing need for tooling that simplifies machine learning operations on Apple silicon. Future updates could expand support for more model types or integrate with other ML pipelines.