Docs→Budget limits

Budget limits & cost protection

Set monthly caps on LLM spending — per workspace and per user. Automatic alerts at 50%, 80% and 100%. Access is blocked when the limit is reached. No surprises on your API bill.

How it works

Custos AI tracks estimated token usage and cost per user in real time. When a user approaches their limit, they receive an automatic alert. When they hit the limit, their access to the chat interface is temporarily blocked until the next billing cycle or until an admin resets the limit.

50%User receives an in-app warning notification
80%User receives a second warning; admin is also notified
100%Chat access is blocked for that user until the next cycle or manual reset by admin

Set a workspace budget

  1. 1Go to Settings → Budget in the Admin dashboard.
  2. 2Enter a monthly limit in euros (estimated cost based on token usage).
  3. 3Click Save. The limit applies to all users in the workspace combined.

Set per-user limits

  1. 1Go to Settings → Team in the Admin dashboard.
  2. 2Click on a team member to open their settings.
  3. 3Enter a monthly limit for that user and click Save.
  4. 4Per-user limits are enforced independently of the workspace limit. Both apply simultaneously.

Important note on cost estimates

LLM costs are estimated based on token usage and published pricing of each provider. Actual costs may differ slightly due to pricing changes or rounding. Custos AI is not liable for overruns — always set limits with a safety margin and monitor your actual spend in your provider account.

Next: Team management

Invite members, assign roles, and control provider access per user.

Team management