AI Inference Admin
AI Inference Admin is the admin app for viewing and managing inference resources across all projects in one place. Launch AI inference from Admin Center on the desktop.
Unlike the user app, you don’t choose a cluster or project — you see all projects at once — and it has Token metering, Kueue, Monitoring, and project management menus that the user app lacks. It cannot create endpoints; create them in AI Inference.
Menus
| Group | Menu | What it does |
|---|---|---|
| — | Dashboard | View overall storage and resource usage, workload and volume summaries, and recent activity. |
| Serverless · vLLM | Docker · GPU · CPU · Rebellions | View and manage endpoints of all projects. |
| Infrastructure | Templates · Workloads · Volumes | View, edit, and delete templates, workloads, and volumes of all projects. |
| Evaluation | Benchmark · Token metering | View benchmark results and token usage by endpoint. |
| Kueue | Monitoring · Queue · Workloads · Analytics · Scheduler · Resource Inventory | Manage GPU queues and scheduling. |
| Monitoring | Nodes · Usage Trend · Connected Clusters | View node status, usage trends, and connected clusters. |
| Admin Managements | Projects | Create and manage projects. |