Rate Limits

Rate limits protect Pixapi and upstream model providers from sudden traffic spikes. Limits may vary by plan, model family, and endpoint.

Common signals

Pixapi may return 429 rate_limited when the current key or account exceeds its allowed request rate.

{
  "error": {
    "code": "rate_limited",
    "message": "Too many requests. Please retry later."
  }
}

Client behavior

• Queue bursty jobs in your backend.
• Use exponential backoff with jitter.
• Avoid retrying media generation without checking task state.
• Split production, staging, and development traffic across separate API keys.

Backoff with jitter

function backoff(attempt: number) {
  const base = Math.min(8000, 500 * 2 ** attempt);
  return base + Math.floor(Math.random() * 250);
}

Increasing limits

If you run high-volume workloads, contact support with your expected request volume, model mix, and peak traffic window.