Licensed to be used in conjunction with basebox, only.
// admin
Dashboard
The admin dashboard shows how much your organization uses basebox: tokens, calls, active users and response times, broken down by app, model and user. It changes no settings apart from the prices and limits you enter there. Open it under Administration → Dashboard; only administrators see it.
Filters decide which usage you see
The filter bar at the top applies to every section of the page:
- Period – last 7 days, last 30 days, this month, last month, since the start, or a custom range.
- Compare – no comparison or previous period. With a comparison, the key figures show the change against the previous period.
- Org unit, App and Model – narrow the analysis.
- Channel – Web app, API, Workflows or speech recognition.
- Group by – day, week or month for the chart.
Export CSV at the top right downloads the usage for the selected filters as a CSV file.
Five key figures summarize usage
- Tokens – tokens used in the period.
- Calls – requests to the AI models.
- Avg tokens per call – high values point to long conversations or large attachments.
- Active users – how many registered users used basebox in the period.
- Time to first token – how long it takes until an answer starts (median and 95th percentile).
Needs attention lists anomalies
The Needs attention section lists warnings, for example requests with a very large number of tokens or unusual access (several client addresses per person, many calls at night or at weekends). Show jumps to the details; the list can be exported as CSV.
Chart and tables break usage down
- Tokens and calls over time – bars for tokens, a line for calls; dashed lines mark basebox releases.
- By app and By model – tokens, share and calls per app or model.
- How usage is spread across users – share of the heaviest user and the ten heaviest users, active users, usage per active user and users without usage.
- Largest conversations and Request size – show where single requests become very large.
- Access and abuse check – client addresses from the audit log.
- When people work – usage by time of day and weekday.
You set prices and limits in the dashboard
Models and pricing shows the Monthly limit in tokens, the usage in the current month and the date on which the limit resets.
- Prices – if you enter prices per model for input, output and cached tokens, the dashboard shows estimated costs. Without prices it shows no costs. The estimate is not checked against an invoice.
- Limits – if you set thresholds, the dashboard warns under Needs attention as soon as they are exceeded.
Frequently asked questions
Who can see the dashboard? Only administrators of your organization.
What are tokens? Tokens are the unit in which use of the AI models is measured. In English, a word is roughly one to two tokens.
Is API usage shown separately? Yes. Choose API in the Channel filter.
Next step: Managing apps
Need help? Contact support