Error Codes

Stable HTTP status and business error code mappings used by gpt88.cc for client-side retry and troubleshooting logic.

Error envelope

error envelopejson
{
  "error": {
    "type": "rate_limit_error",
    "code": "rate_limit_exceeded",
    "message": "Rate limit exceeded for model gpt-5.6-sol",
    "param": null,
    "request_id": "req_01HZX3..."
  }
}

Common codes

HTTPcodemeaning
400invalid_request_errorMalformed request body or invalid field shape.
401invalid_api_keyMissing, revoked, or invalid API key.
403permission_deniedThe current key cannot access the requested model or endpoint.
404model_not_foundModel ID does not exist or is not available to the key.
408request_timeoutThe upstream response exceeded the gateway timeout window.
429rate_limit_exceededRate limit was hit at account, key, or model level.
429insufficient_quotaBalance or quota is exhausted.
500internal_errorGateway or upstream internal failure.
502upstream_errorUnexpected upstream provider response.
503service_unavailableUpstream capacity is temporarily unavailable.
504gateway_timeoutGateway timed out while waiting for upstream.

Retry guidance