Usage & Limits

Usage & Limits

Rate Limits

Each tier has different rate limits:

Tier	Requests/Minute	Monthly Limit
Free	10	300 (~10/day)
Lite	60	1,000
Pro	600	Unlimited*

*Unlimited subject to fair use.

Rate Limit Headers

All API responses include rate limit information:

X-RateLimit-Limit: 60
X-RateLimit-Remaining: 45
X-RateLimit-Reset: 1718169660

Checking Usage

curl https://cryptgpt.co/billing/usage \
  -H "Authorization: Bearer YOUR_API_KEY"

Response:

{
  "tier": "pro",
  "monthly_usage": 1234,
  "monthly_limit": null,
  "rate_limit_rpm": 600,
  "alert": null
}

Usage Alerts

When approaching your tier limit, the alert field will indicate:

Alert	Threshold	Meaning
null	< 80%	Normal usage
warning	80-89%	Consider upgrading
critical	90-99%	Approaching limit
exceeded	100%	Limit reached

When You Exceed Limits

Rate Limit (429)

If you exceed requests per minute:

{
  "error": {
    "message": "Rate limit exceeded",
    "type": "rate_limit_error",
    "code": null
  }
}

Wait for the Retry-After header duration, then retry.

Monthly Limit

If you exceed your monthly message limit:

{
  "error": {
    "message": "Rate limit exceeded",
    "type": "rate_limit_error",
    "code": null
  }
}

Upgrade your tier or wait for the next billing cycle.

Best Practices

1. Cache responses when possible
2. Use streaming for long responses
3. Implement exponential backoff when retrying
4. Monitor remaining requests via headers
5. Batch requests when possible