Control usage and costs by setting token limits per user or per team through the Nexus portal.
Quota policies let you define spending and usage boundaries for your organization. Policies are managed entirely through the Nexus portal web interface — no CLI or configuration files required.
| Field | Description |
|---|---|
| Policy Name | A descriptive name (e.g., "Engineering Team Standard") |
| Scope | Per-user or per-team |
| Token Limit | Maximum tokens per period (input + output combined) |
| Period | Daily, weekly, or monthly reset cycle |
| Enforcement Mode | Soft limit (warn) or hard limit (block) |
| Model Restriction | Optionally restrict to specific Claude models |
| Mode | Behavior |
|---|---|
| Soft (Warn) | Users receive a warning when approaching the limit but can continue. Usage is flagged in the Audit Log. |
| Hard (Block) | Requests are denied once the limit is reached. Users see a clear message explaining the quota has been exceeded. |
When a user belongs to a team with a policy, the more restrictive limit applies (individual or team).
Deleting a policy removes it from all assigned users and teams. Those users will have unrestricted access until a new policy is assigned.
You can set a default policy that applies to all users who don't have an explicit assignment:
Conservative (cost control):
Standard (balanced):
Power User (unrestricted):