arnie-ai-granite

Description

This container image provides a self-contained IBM Granite AI model server for the ARNIE AI infrastructure operator. It bundles the Granite 3.1 Dense (8B) model together with the Ollama inference engine on a Red Hat Universal Base Image, so the model runs locally with no external dependencies or runtime downloads. Because inference happens entirely within the deployment environment, no data leaves the customer's infrastructure — making it suitable for security-conscious and air-gapped deployments. The image runs as a non-root user and exposes the model API on port 11434 for use by the ARNIE AI operator. A BLCKBX product.