Usage & Billing

Quotas and usage statistics are available in the Console and via the response payload of each request.

Usage in Response

Every non-streaming response includes a usage object:

"usage": {
  "prompt_tokens": 120,
  "completion_tokens": 45,
  "total_tokens": 165
}

Quota

Quota is measured in tokens. Admins set per-user quotas from Console → Members. Users can monitor their remaining quota in Console → Limits & Billing.

ℹ

Streaming responses do not include a usage field in intermediate chunks. A final synthetic usage summary is available in the last chunk before data: [DONE].

Turning token counts into a budget? Feed your monthly volumes into the LLM API cost calculator, or check per-token rates in the model price comparison.