Skip to content

Govern AI spend

You govern AI spend from two places in the admin sidebar’s Usage & Control group: the Control Center, which sets the rules, and the Usage page, which shows where credits went. This page is the map; each lever links to its deep dive. What credits are and what happens at depletion is covered in Credits and AI spend.

The Control Center (“govern credit spend, model defaults, and build-mode access across your organization”) has five tabs, and every edit autosaves:

  • Overview — read-only: where the credit pool stands this period, with a consumption meter and your threshold markers.
  • Allocations — a monthly per-user credit cap (Limited / Unlimited / Blocked), folding org default → role → individual.
  • Build access — which roles and people can enter Build mode, “the most credit-intensive work.”
  • Automation (spend thresholds — unrelated to the Automations feature) — thresholds at a percentage of the pool: Org warning, User warning, and exactly one User downshift (which swaps everyone to the Reduced models).
  • Models & effort — the Standard and Reduced model sets per mode, plus who may pick their own model and effort in chat.

For the full walkthrough of each tab, see Control Center.

Topping up the wallet happens on the separate Billing page and requires the Super Admin: enter a Credit amount, check the live cost preview, and confirm. Credits land immediately; “the cost is added as a line on your next invoice — your card is not charged.” The full flow is in Usage & billing.

The Usage page traces every credit from the source that triggered it to the model behind it, and lets you group spend by user, model, or mode. See Usage & billing.