API
The Gcore Customer Portal is being updated. Screenshots may not show the current version.
Edge AI
Edge AI
Billing
About GPU CloudCreate an AI Cluster
Prepare a custom model for deployment
Deploy AI models in the Customer PortalDeploy AI models in the Customer Portal
Manage deployments in the Customer PortalCreate and manage API keysCreate and configure a registry
API
Chosen image
Home/Edge AI

Add a container image registry

Everywhere Inference supports custom private and public registries in addition to Gcore ones, so you can manage your model images with any service you like.

Step 1. Initialize a new registry addition

You can initialize a new registry by following these two steps.

1. In the Gcore Customer Portal, navigate to Everywhere Inference > Registries.

Add a registry

2. Click Add registry in the top right of the screen.

Add a registry

This will open the Add registry dialog, where you can add the authentication credentials for the registry you want to add.

Step 2. Enter registry configuration

Enter the registry’s name, URL, username, and password.

The URL for the Huggingface registry is “”. You can create an access token in the Huggingface settings. Use “username” as the Image registry username and the access token as the Image registry password. Add a registry

Step 3. Finalize registry addition

After you enter all registry details, click Add. Your new registry’s name should appear in the list of registries.

You can now select the registry for a custom AI model deployment. To learn how to deploy a custom AI model, read our guide.

Was this article helpful?