Rate limits & quotas

Two independent controls protect the platform: rate limits (requests per minute) and token quotas (total tokens per window). Both are configured per plan and enforced server-side.

Rate limits

Each API key is limited to a requests-per-minute ceiling defined by your plan. Exceeding it returns 429 with a rate_limit_error and a Retry-After header.

Token quotas

Your plan includes a token allowance per window (daily or monthly). Usage is metered on the actual tokens returned by the upstream model; when a provider doesn't report usage, a documented estimator is used and the record is flagged as estimated.

{
  "error": {
    "message": "Token quota exceeded. Used 30,000,000 of 30,000,000 for this day.",
    "type": "quota_error",
    "code": "quota_exceeded",
    "request_id": "req_..."
  }
}

Plan allowances

• Free — 5M tokens / month
• Starter — 30M tokens / day
• Growth — 100M tokens / day
• Scale — 500M tokens / day

Allowances are configurable and may change; your dashboard always shows your live quota and remaining balance.