Rate limits & fair use

The requests-per-minute budgets the edge actually enforces, and what a 429 tells you.

Per-key limits

Each API key gets a requests-per-minute budget from its plan tier, measured over a sliding 60-second window. Exceeding it returns 429 with a Retry-After header (seconds until the window frees up) — wait that long or upgrade for a higher cap.

Plan	Requests / minute per key
Free	10
Intelliverse Air	60
Intelliverse Pro	300
Intelliverse Pro Max	1000

Important

These are the numbers the edge actually enforces, not aspirational ones — a drift-guard test fails the build if this table ever diverges from enforcement.

Shared-IP ceiling

Independently of the per-key cap, one client IP is limited to 120 requests/minute across all keys (NAT / office egress guard). The 429 body says which limit you hit: key_rate_limit or ip_rate_limit. If a shared office IP caps your team, contact us about dedicated throughput.

Spend-velocity review

Spending more than 500 credits inside 5 minutes on one key flags the traffic for review. It never blocks the request — it is an abuse signal, not a limit.

Anatomy of a 429

HTTP

HTTP/1.1 429 Too Many Requests
Retry-After: 42

{
  "error": {
    "code": 429,
    "message": "Rate limit exceeded. Slow down or upgrade your plan for a higher RPM cap.",
    "metadata": { "reason": "key_rate_limit", "retry_after": 42 }
  }
}

Honor Retry-After with exponential backoff as a fallback. The full error catalog is in Errors.