Admin ConsoleMonitor

Usage analytics

See your organization's spend, requests, tokens and latency, break them down by model, person, provider and team, and email yourself a report.

Note
Required role
Owner or Admin of the active organization. Everyone can see their own usage on My Usage.

All figures cover the active organization only, and costs are in US dollars at the prices set in Settings → Model Settings. They count requests over a date range, so they can differ from live credit balances.

Overview dashboard

Overview is the first entry in the Administration group: a summary of the organization at a glance. Pick the range at the top right: 7d, 30d (the default), 90d or All.

  • Spend, Requests, Input tokens and Output tokens for the range.
  • Users, Teams and Tiers counts; select one to open that page.
  • Spend over time and Requests over time charts.
  • Top models by spend and Top users by spend; select a user to see their usage. Analytics → opens Usage Analytics.

Usage Analytics

Open Usage Analytics from the Administration group for the full breakdown.

Filter

  1. 01
    Select Filters

    A badge on the button counts the filters in use.

  2. 02
    Choose a Time frame

    Today, 7d (the default), 30d, 90d, 1yr, All, or Custom for a start and end date.

  3. 03
    Narrow it down

    Optionally pick one Model, User or Product (for example LatentCode, LatentWork, or direct API calls).

  4. 04
    Select Apply

    Every chart and table on the page updates.

What the page shows

SectionWhat it shows
Summary cardsTotal Cost, Total Requests, Input Tokens, Output Tokens, Avg Cost / Req, Cache Hit Rate, and average input, output, cache-read and cache-write cost per request.
Token breakdown (cost drivers)How input, output and cache tokens split, to show what drives cost.
Requests heatmapRequests by day of week and hour, in India Standard Time (IST). Hover a cell for that hour, or a day label for the whole day.
Daily Cost, Daily RequestsDay-by-day totals.
Cost Trend by ProviderDaily cost, one line per provider.
Cost by Model, Requests by Model, Avg Latency by ModelPer-model comparisons.
Latency DistributionHow request latencies are spread.
Cost by ProviderShare of cost per provider.
Top Users by CostThe heaviest users, with requests, cost and the teams they used.
Usage ExplorerA table you can Group by Model, User, Provider or Project (team), with extra provider and team filters. Columns: requests, cost, input, output, cache-read and cache-write tokens. Grouped by user, select a row for that person's full analytics.

Email a report

Select Email report to send a summary for the selected time frame to your own email address. The page confirms where it was sent. The report covers the whole organization; the model, user and product filters aren't applied to it.

One person's usage

On the Users page, select ⋮ then View usage on the person's row. You'll see their spend, requests, per-model table and charts in this organization, plus their tier's allowed models, rate limits and budgets with how much of each they've used. It's the admin's view of the person's My Usage page.

Troubleshooting

The dashboard is empty

There's no usage in the selected range for the active organization. Widen the range, and check the organization switcher.

Totals don't match a credit balance

Analytics count requests in a date range; a credit balance counts spending since the limit was last set or reset.

I was sent to My Usage instead of the dashboard

The dashboard needs the Owner or Admin role in the active organization.