Errors & rate limits

ZekronAI uses conventional HTTP status codes and a consistent JSON error shape so failures are easy to handle programmatically.

Error shape

json

{
  "error": {
    "type": "insufficient_balance",
    "message": "Your balance is too low to start this task.",
    "code": 402
  }
}

HTTP	Type	Meaning
400	invalid_request	The request body was malformed or a required input field was missing.
401	authentication_error	The API key is missing, malformed, or has been revoked.
402	insufficient_balance	Your account balance cannot cover the estimated task cost. Top up to continue.
403	permission_denied	The key is not allowed to use this model category.
404	not_found	The task ID or model does not exist.
422	unsupported_parameter	An input parameter is not supported by the chosen model.
429	rate_limited	Too many requests. Slow down and retry after the indicated delay.
500	server_error	An unexpected error occurred. Failed tasks are never charged.

Limits are applied per account and returned in the X-RateLimit-Remaining response header.
When you exceed a limit you receive a 429 with a Retry-After header — back off and retry.
Need higher throughput? Request a limit increase from your dashboard or support.