Managing API budgets for a team
Organization credits, per-member caps, per-key budgets and token quotas.
When several people share one payment method for LLM APIs, two things go wrong: nobody knows who spent what, and one person's mistake can burn the team's balance. PleumRouter separates those concerns by layering per-member and per-key caps on top of organization credits. Caps are checked before the upfront hold, so an over-limit request never goes out and gets reconciled later.
What does an Organization share?
Credits and usage aggregation. Create an organization, invite members, top up once, and everyone calls with their own API key. There are four roles — owner, admin, member and viewer — and a viewer can see usage but cannot create keys.
How do I stop one member from draining the team balance?
Set a per-member monthly budget. It aggregates usage across every API key that member holds, so creating extra keys does not work around it. Leave it empty for unlimited.
What limits can I put on a single API key?
Daily, weekly and monthly budgets, plus a monthly token quota (input and output combined) that is independent of the money budget. On top of that you can restrict the allowed model list and requests per minute. A practical split is a generous budget on production keys and a short daily budget on development and experiment keys.
What happens when a limit is exceeded?
The call is rejected and not billed. Exceeding the token quota returns 402 with monthly_token_quota_exceeded. The check happens immediately before the upfront hold, so an over-limit request never reaches the provider.
If a key leaks, how far does the damage go?
As far as the limits you set in advance. A tight daily budget and a narrow allowed-model list bound the loss until you notice. Deactivate the key from the dashboard as soon as you do.
How do I see who spent what?
The usage screen in the dashboard aggregates by organization, member and key, with a breakdown by model and period, so you can isolate which team and which workload is driving cost.
What applies where
| Control | Scope |
|---|---|
| Member monthly budget | All keys held by that member, combined |
| Key daily / weekly / monthly budget | A single key |
| Monthly token quota | A single key (input + output, independent of money) |
| Allowed models | A single key |
| Requests per minute | A single key |
Last reviewed: 2026-09-02 · /docs/limits · /docs/api/keys