For the complete documentation index, see llms.txt. Markdown versions of all docs pages are available by appending .md to any docs URL.

Cost dashboard

View LLM spend, tokens, and traffic in the built-in Admin UI when running agentgateway on Kubernetes.

The built-in Admin UI includes a cost dashboard: the same LLM > Analytics page available in standalone mode. It plots spend, tokens, and calls over time, broken down by model, provider, user, group, or user agent. It works in Kubernetes too, but unlike standalone it is not enabled by default. The proxy ships without a request-log database, so the dashboard has nothing to show until you configure one.

Requirements

Two pieces of configuration power the dashboard:

• A request-log database (config.database). The Kubernetes AgentgatewayParameters resource has no typed field for it, so you set it through rawConfig, which merges raw agentgateway configuration into the proxy’s config file. Point it at a writable path in the pod, such as the /tmp volume.
• A model cost catalog (spec.modelCatalog) so requests are priced. Without a catalog, the dashboard still shows token and call volume, but the cost is 0.

Enable the dashboard

1. Create (or update) a Gateway-level AgentgatewayParameters resource that enables the database through rawConfig and attaches a model catalog. The catalog configMap must already exist in the Gateway’s namespace (see Model costs).

   kubectl apply -f- <<EOF
   apiVersion: agentgateway.dev/v1alpha1
   kind: AgentgatewayParameters
   metadata:
     name: cost-dashboard-params
     namespace: agentgateway-system
   spec:
     modelCatalog:
       sources:
         - configMap:
             name: my-model-costs
             key: catalog.json
     rawConfig:
       config:
         database:
           url: "sqlite:////tmp/data.db?mode=rwc"
   EOF

2. Attach the AgentgatewayParameters resource to your Gateway with infrastructure.parametersRef. modelCatalog and rawConfig are honored only on a Gateway-level resource.

   kubectl apply -f- <<EOF
   apiVersion: gateway.networking.k8s.io/v1
   kind: Gateway
   metadata:
     name: agentgateway-proxy
     namespace: agentgateway-system
   spec:
     gatewayClassName: agentgateway
     infrastructure:
       parametersRef:
         name: cost-dashboard-params
         group: agentgateway.dev
         kind: AgentgatewayParameters
     listeners:
       - name: http
         port: 80
         protocol: HTTP
         allowedRoutes:
           namespaces:
             from: All
   EOF

3. Wait for the proxy to roll out with the new configuration.

   kubectl rollout status deployment/agentgateway-proxy -n agentgateway-system

Open the dashboard

Every request that flows through the proxy is now recorded and priced. Port-forward the proxy’s Admin UI and open the Analytics page.

1. Port-forward the Admin UI port.

   kubectl port-forward deployment/agentgateway-proxy -n agentgateway-system 15000

2. Open http://localhost:15000/ui/llm/analytics. Send some LLM traffic through the gateway, then refresh to see traffic over time with a running tally of cost, tokens, and calls, plus a breakdown below the chart.

   

   

Group by and measure

Use the Group by control to break the same traffic down by model, provider, user, group, or user agent*, and toggle Measure between tokens, cost, and requests. Set to Cost to see realized spend in dollars. Use Export to pull the underlying numbers out for reporting.