Nucleus is a minimalist, security-hardened container runtime designed for Linux. It features a fully declarative service model built on NixOS, allowing precise and reproducible configuration management [1]. It supports three operating modes: agent mode (the default), strict agent mode, and production mode, which targets long-running NixOS services [1]. Nucleus delivers near bare-metal performance specifically for PostgreSQL workloads running on Linux 6.18 x86_64 kernels, offering efficiency for database applications [1].
Meanwhile, Apple introduced a lightweight container runtime called container, optimized for Apple silicon Macs running macOS 26 [2]. This runtime supports OCI-compatible container images and manages containers, images, and processes with a Swift package [2]. The tool requires macOS 26 and Apple silicon hardware, and it includes scripts to install, upgrade, downgrade, or uninstall the runtime, simplifying lifecycle management for users [2].
Apple’s container machine offers a highly integrated and persistent Linux environment on Macs. It enables seamless sharing of user and home directories between macOS and the Linux container [3]. Users can edit code on macOS with compilation and execution performed inside Linux container environments that match various Linux distributions [3]. Furthermore, container machine supports running full Linux system services like PostgreSQL using systemctl inside containerized Linux virtual machines on Mac [3].
The Apple container runtime becomes available following the release of macOS 26, which is a requirement for its operation [2]. Both Nucleus and Apple’s container tool cater to different segments: Nucleus focuses on secure, high-performance Linux containers with declarative service control, while Apple’s solution targets developers who want integrated Linux environments on macOS with native Apple silicon support.
Developers and system administrators can now evaluate these runtimes based on their platform needs and workload requirements, as both aim to improve container usability and performance in their respective ecosystems.
