Hippius Hub Registry

The AI model registry built for decentralized compute. Hippius Hub is a public registry for AI models and OCI artifacts — pull anything anonymously, push with an account.

It has two faces:

Face	What it is	Who uses it
Model Registry	A drop-in replacement for Hugging Face Hub. Same Python API (hf_hub_download, snapshot_download, from_pretrained), backed by an OCI registry instead of huggingface.co.	ML engineers loading models with transformers, diffusers, or any library that uses huggingface_hub under the hood.
Container Registry	A standard OCI registry. Push and pull Docker images, ORAS artifacts, and raw OCI blobs against registry.hippius.com.	Anyone running docker push / docker pull, plus power users with oras for non-image artifacts.

Both faces share the same authentication, the same backing storage, and the same endpoint.

---

Endpoints

Purpose	URL
OCI registry (Docker, ORAS, hippius-hub)	registry.hippius.com
Web UI (browse models & containers)	hub.hippius.com
Source / issues	github.com/thenervelab/hippius-hub

Models are indexed server-side by format, architecture, parameter count, and quantization — so search on hub.hippius.com works the way you'd expect.

---

Five-minute path

Pick the flow that matches what you're shipping. The CLI install and namespace provisioning are the same either way.

Models (Python)

# 1. Install the unified CLI (also the Python library)
pip install hippius_hub

# 2. Save your API token from https://console.hippius.com/dashboard/settings
hippius-hub login --hippius-token <paste-token-here>

# 3. Provision a namespace + run docker login in one shot
hippius-hub registry provision my-models --docker-login

# 4. Push a model folder as :v1
hippius-hub upload my-models/qwen-7b ./qwen-7b --revision v1

Pulling from Python is a one-line import swap — your existing transformers / diffusers code keeps working:

import hippius_hub as huggingface_hub
from transformers import AutoModel

model = AutoModel.from_pretrained("my-models/qwen-7b")

Containers (Docker)

# 1. Install the CLI for namespace provisioning + credential helpers
pip install hippius_hub

# 2. Save your API token from https://console.hippius.com/dashboard/settings
hippius-hub login --hippius-token <paste-token-here>

# 3. Provision a namespace and run docker login in one shot
hippius-hub registry provision my-models --docker-login

# 4. Tag and push an image
docker tag my-app:latest registry.hippius.com/my-models/my-app:v1
docker push registry.hippius.com/my-models/my-app:v1

Pulling a public image needs no login:

docker pull registry.hippius.com/my-models/my-app:v1

Same Python API as Hugging Face

The Model Registry is a drop-in for huggingface_hub. Same cache layout, same function signatures, same exception classes — so transformers.from_pretrained(...) and any other code that uses huggingface_hub under the hood works without modification. See Pull for the full Python surface.

---

Hub access from the Console

Browse, manage, and inspect Hub repos directly from console.hippius.com.

Repositories — see every container and model you've pushed, with pull counts and last-modified timestamps.

Manage — pick a subscription plan, toggle namespace public/private access, and rotate docker credentials for CI/CD.

Pricing details are available at hippius.com/hippius-hub.

---

Where to next

• Pull — download a single file or a whole repo from Python, the CLI, or docker pull.
• Push — provision a namespace, push artifacts, manage credentials.
• CLI reference — the hippius-hub command surface, grouped by goal.