/organizations/:name/insights/usage). Organization admins see the whole organization; other members get a read-only view of their own usage.
Users tab
The Users tab is available to organization admins on Usage > Overview, one of its breakdown tabs — Workspaces, Agents, Users, Triggers, Platforms, Providers, and Models. The dashboard opens on Workspaces. It attributes LLM spend to the accountable human for each session, resolved as the chat initiator, then the test-session initiator, then the creator of the trigger that fired the session. Each row represents one organization member. When no accountable human can be resolved, that spend collects into a single Automated (no user) remainder row rather than disappearing from the table. The remainder row is not clickable and always sorts last, which keeps the tab reconciled to the organization total — so the % of spend column sums to 100%. Expand the Automated (no user) row for an in-place drill-down of what drove the automated spend. It groups by the firing trigger, falls back to the agent name where there is no trigger, and collects anything with neither under an Other (no trigger or agent) remainder. Each breakdown tab shows a % attributed indicator in the top right, reporting how much of the total is explained by named rows rather than the remainder.Usage forecast
The forecast chart projects future spend from your recent run rate, drawing a dashed projection against your budget line so you can see whether current activity trends over budget.What if slider
Use the What if slider below the chart to model a change in activity. It sets a run-rate multiplier from 0.5× to 10× in 0.5× steps, rescaling the forward projection in place against the budget line. A readout pairs the multiplier with the implied monthly cost, in the form2× current · ~$480/mo.
The slider scales only the forward projection. Spend you have already incurred is not rescaled.