Skip to main content

Overview

This page summarizes the most common deployment actions available from project deployment views.

Action Flow

Core Actions

ActionWherePurpose
Deploy ModelDeployments listCreate a new endpoint from selected model + cluster
Use this modelList row or General tabCopy cURL, Python, and JavaScript snippets for immediate endpoint use
PublishDeployments list rowPublish model to the Customer Dashboard portal with pricing metadata
Publish DetailsPublished endpoint rowReview or adjust publication pricing metadata
UnpublishPublished endpoint rowRemove model from customer portal while keeping deployment active
DeleteDeployments list rowRemove endpoint after dependency checks

Detail-Page Actions

TabActionUse Case
GeneralUse this model / View deployment summaryAccess endpoint snippets and validate model-cluster-status context
WorkersAdd WorkerIncrease local serving capacity
SettingsSave policy changesApply rate limits, retries, and fallback rules

Operational Notes

  • Deploying a model makes the endpoint usable; publishing is not required for internal API use.
  • Publish/unpublish controls specifically govern visibility in the customer-facing portal.
  • Workers tab appears only for non-cloud deployments.