Cloudera Model Hub

A curated catalog of NVIDIA NIM microservices — production-ready AI models spanning language, vision, speech, and more — optimized for deployment with Cloudera AI Inference Service.

Browse the catalog

→ Open the Model Hub catalog

Models can be imported directly into the Cloudera AI Registry and deployed with a single click via Cloudera AI Inference Service.

Repository structure

models/public/          # Model descriptor YAMLs (latest)
models/private/         # Private cloud variants
models-order/public/    # Merge order defining catalog sequence
manifests/              # Model Registry manifest versions
airgap-scripts/         # Scripts for airgapped deployments

Adding or updating a model

1. Add or edit the model YAML in models/public/ following the schema in utils/base_model.yaml.
2. Add the filename to models-order/public/merge-order.yaml in the desired position.
3. Commit to main — the catalog page picks up changes immediately, no build step required.

GitHub Pages setup

The catalog is served directly from index.html at the repo root. It fetches the YAML files at runtime.

Internal operations

See OPERATIONS.md for Jenkins jobs used to sync content into mlx-crud-app and model-registry.