ModelPack - An Open Standard for Packaging, Distributing and Running LLMs in Cloud Native Environment

Explore ModelPack, an open standard for packaging, distributing, and running AI artifacts in cloud-native environments through this 27-minute conference talk from the Linux Foundation. Discover how infrastructure is evolving from container-centric to model-centric approaches, with ModelPack serving as the bridge between traditional cloud-native infrastructure and modern AI/LLM workloads. Learn about the technical specifications that standardize AI/LLM models as OCI artifacts, connecting leading open-source cloud-native projects to create seamless integration. Examine the current community status of this open standard and witness practical implementations at Ant Group, demonstrating how ModelPack enables efficient management and operation of AI/LLM models using existing cloud-native infrastructure including Kubernetes and container runtimes. Understand how this specification addresses the growing need for standardized AI model distribution and deployment in enterprise environments, providing a foundation for the next generation of model-centric infrastructure development.