AI Usage
AI Usage shows how much AI activity your project records: how many traces, observations and scores arrive, how that volume changes over time, and where it comes from. Use it to understand load and ingestion volume. For spend, use the Cost Dashboard.
Open it from Gauge AI → AI Usage. It opens the prebuilt Usage Management dashboard and needs Gauge AI read access (or the Cost & Latency dashboards sub-module): Owner, Department Admin, Security & Compliance and Cost Analyst have it.
What it shows
| Widget | What it shows |
|---|---|
| Total Trace Count and Total Observation Count | Totals for the selected range. |
| Total Score Count (numeric) and (categorical) | How many scores were recorded, for example by evaluators. |
| Total Trace Count (over time) and Total Observation Count (over time) | Volume trends. |
| Total Score Count over time | Numeric and categorical scores per period. |
| Total Trace Count (by env) and Total Observation Count (by env) | Volume per environment, such as production and staging. |
| Usage by Gateway Model Group | Total tokens per Budget AI model group. |
Filters
The dashboard defaults to the last 7 days, and the range you can pick is limited by your plan's data access period. The filter builder narrows by environment, trace name, observation name, score name, tags, source, gateway key, gateway model group, user, session, metadata, release and version.
How to use it
- Check the totals and trends for the period to spot growth or sudden drops in traffic.
- Compare environments to confirm that test traffic is not mixed into production.
- If volume rises, open the Cost Dashboard to see what it costs, and Cost & Performance Metrics for tokens and cost per model.
- For a costly agent, run the AI Optimizer.
Ingested traces, observations and scores also count toward credits on credit-based plans. See Credits.
Next steps
- Cost & Latency Dashboards: spend and latency breakdowns.
- Custom Dashboards: clone this dashboard and add your own widgets.