Skip to main content
Edge Inference Services is not yet available in production. This article is published in advance so it is ready when the feature ships.

​
Introduction

The Inference Dashboard is the home page for Edge Inference Services. It provides a real-time summary of your inference environment, including model counts, deployment health, benchmark results, and recent activity. Navigate to it at Inference > Dashboard.

​
Summary Tiles

The top row of the dashboard shows five summary tiles. Each tile displays a live count for that resource type and opens the corresponding page when selected.

​
Quick Actions

The Quick Actions bar provides one-click shortcuts for common tasks. The following actions are available.
  • Add Model opens the form to register a new model in your inference environment.
  • Add Provider opens the form to connect an external model source.
  • Add Benchmark opens the form to create a new benchmark evaluation run.
  • Add Deployment opens the new deployment form.

​
Status Charts

Four charts show the current distribution of resource states across your inference environment. Benchmark Status shows the count of benchmarks in each state (Pending, Queued, Scheduled, Deploying, Running, Completed, Failed, or Cancelled). Deployment Status shows the count of deployment requests in each state (Pending, Approved, Rejected, Closed, Pending Update, or Pending Delete). Import Status shows the state of active model import operations. Sources Health shows the health status of connected model sources (providers). Use these charts to spot failures or stalled operations before they affect production workloads.

​
Models by Status and Framework

The Models by Status & Framework chart breaks down your model inventory by run state and ML framework. Use this chart to see how many models are active versus idle, and which frameworks are in use across your environment.

​
Recent Activity

The bottom of the dashboard shows two recent activity lists. Recent Deployments lists the most recently created or updated deployments. Each row shows the deployment name and the model it runs. Recent Evaluations lists the most recently completed benchmark evaluations. Each row shows the benchmark name and the evaluated model. Select any row to open the full detail page for that item.

​
In This Section