Live monitor

View: Live monitor The live monitor showing the activity sparkline, period cards, and controls.

The Live Monitor view provides a real-time view of gateway activity. It polls the Stats API on a configurable interval and displays continuously updating metrics for the current minute, last hour, and today.

Required role: platform admin. Both the view (/monitor) and its /monitor/stats feed are platform-admin-only — the cross-tenant activity view is not exposed to tenant admins or other roles.

Reach it via Overview › Analytics › Live Monitor in the left rail.

Controls

Control	Description
Currency	Sets the currency used for every cost figure on the page. Costs are metered in USD; EUR is shown as an approximate conversion at the daily reference rate (the last known rate, or a built-in reference rate, is used if the rate service is briefly unavailable).
Tenant filter	Scopes all metrics to a single tenant. Defaults to all tenants.
Interval	Auto-refresh interval: 1 s, 2 s, 3 s (default), 5 s, 10 s, or 30 s.
Pause / Resume	Stops or restarts auto-refresh. Metrics freeze on pause.
Refresh	Forces an immediate update regardless of the current interval.

The header shows the timestamp of the last successful poll and whether auto-refresh is running.

Activity sparkline

A rolling line chart plots the trailing-60-second request count (labelled Requests/min) over the last 60 samples. The current value appears to the right of the chart.

Period cards

Three cards break down activity across time windows:

Card	Window
Last minute	Rolling 60-second window
Last hour	Rolling 60-minute window
Today	From your local calendar midnight to now

Each card shows the following metrics:

Metric	Description
Requests	Total inference requests
Cached	Requests served from cache
Blocked	Requests blocked by guardrails or budget
Scrubbed	Requests where a guardrail scrubbed content (action = `scrub`)
Flagged	Requests flagged by a guardrail (action = `flag`)
Cost	Total estimated cost in USD
Saved	Cost saved by cache hits
Avg latency	Average end-to-end latency in milliseconds
Provider ms	Average upstream (provider) latency in milliseconds
Input tokens	Total prompt tokens consumed
Output tokens	Total completion tokens generated

Tenants — Today

The Tenants — Today table breaks down today's activity per tenant. It is visible only when at least one tenant has activity.

Column	Description
Tenant	Tenant slug
Requests	Total requests today
Input Tokens	Total prompt tokens consumed
Output Tokens	Total completion tokens generated
Cost	Total estimated cost
Quota Left	The tenant's remaining budget for its current budget period (effective budget minus current-period spend), or no limit when the tenant has no budget. Shows — when the tenant has a budget but its spend could not be read momentarily. An over-cap tenant shows a negative value.

Last 10 Requests

The Last 10 Requests table shows the ten most recent requests across the selected scope.

Column	Description
Time	Timestamp
Tenant	Tenant slug
Provider	Provider that handled the request
Model	Model name
Status	HTTP status code badge, or a blocked / cached badge where applicable
In	Input token count
Out	Output token count
Cost	Estimated cost
Saved	Cost saved by a cache hit, or `—`
Latency	End-to-end latency in milliseconds
Provider ms	Upstream (provider) latency in milliseconds, or `—`

Recent guardrail events

The Recent Guardrail Events table shows the most recent blocked requests. It is like the one on the Dashboard but adds an extra Guardrail (evaluation-latency) column the Dashboard does not show. It is visible only when blocked events exist for the selected scope.