Rate limits

Per-key RPM limits, headers, backoff strategy, and how to raise limits.

Rimp enforces a sliding-window rate limit per API key. Limits are tied to your plan.

Default limits

Plan	Requests / minute	Concurrent jobs
Free	60	1
Pro	300	10
Studio	600	25
Team	1,200	30

You can have multiple API keys per organization, each with its own rate-limit bucket.

Headers

Every response includes:

X-RateLimit-Limit: 300
X-RateLimit-Remaining: 287
X-RateLimit-Reset: 1715900400
X-Request-Id: req_abc

X-RateLimit-Reset is a Unix timestamp (seconds) — wait until then to be safe.

429 response

HTTP/1.1 429 Too Many Requests
Retry-After: 12
X-RateLimit-Limit: 300
X-RateLimit-Remaining: 0
X-RateLimit-Reset: 1715900412

{
  "error": {
    "type": "rate_limited",
    "code": "rate_limited",
    "message": "Rate limit exceeded for this API key (300 rpm)",
    "request_id": "req_abc"
  }
}

Honor Retry-After (seconds). It’s always smaller than 60 for RPM limits.

Recommended backoff

async function withRateLimitBackoff<T>(fn: () => Promise<T>): Promise<T> {
  while (true) {
    try {
      return await fn();
    } catch (err) {
      if (err.status !== 429) throw err;
      const retryAfter = Number(err.headers?.['retry-after'] ?? 1);
      await new Promise((r) => setTimeout(r, retryAfter * 1000));
    }
  }
}

Anomaly detection

If your spending in any hour exceeds 10× your weekly hourly average, Rimp pauses new requests for 30 minutes and emits a usage.anomaly_detected webhook to your account. This protects against credential leaks and runaway loops.

You’ll see:

HTTP/1.1 429 Too Many Requests
Retry-After: 1800
X-Request-Id: req_xyz

{
  "error": {
    "type": "rate_limited",
    "code": "anomaly_paused",
    "message": "Org temporarily paused due to spending anomaly. Resumes at 2026-05-19T16:30:00Z."
  }
}

To resume immediately, contact [e-mail] with your request_id.

Raising your limits

• Pro / Studio: limits raise automatically when you upgrade in /billing.
• Team: write to [e-mail] with your expected RPM and concurrency.
• Burst handling: we don’t charge for unused capacity — burst-bias your traffic into evening windows if practical.