Updated on 2025-07-01 GMT+08:00

Error Codes

If an error code starting with APIGW is returned after you call an API, resolve the problem by referring to Error Codes. If an error code starts with APIG, rectify the fault by referring to this document.

Table 1 Error Codes

Module

Error Code

Error Message

Description

Solution

Model inference

PANGU.0010

parameter illegal.

The request parameter is incorrect.

Enter correct request parameters by referring to the API document and debug the API again.

PANGU.0011

Authentication failed.

Authentication failed.

Authentication failed. For details, see Authentication in the API document.

PANGU.0012

The authentication information is missing.

Identity authentication information is unavailable.

Check whether the authentication information is provided when the API is called.

PANGU.0031

Inner service exception.

Internal error.

Contact technical support for assistance.

PANGU.3254

The requested inference service does not exist.

The resource does not exist.

Check whether projectId and deploymentId are correctly set when the API is called and whether the inference service is available.

PANGU.3267

The number of service invoking requests exceeds the project limit.

Too frequent API requests.

Reduce your request frequency.

PANGU.3278

required api parameter is not present.

The request parameter is lost.

Check whether the request parameters are complete, whether the spelling is correct, and whether the values are correct.

PANGU.3318

The total length of the question should be between 1 and 4096.

The length of Content is invalid.

Check whether the length of the Content parameter value in the request is within the allowed range by referring to the API document and debug the API again.

PANGU.3320

The parameter [n] can only be 1 or 2 when calling non-streaming.

The value of n used for a non-streaming call of the inference service must be 1 or 2.

Use the correct value 1 or 2.

PANGU.3321

The parameter [n] can only be 1 when calling streaming.

The value of n used for a streaming call of the inference service must be 1.

Use the correct value 1.

PANGU.3342

Failed to invoke the inference service. please check the details field.

Failed to call the inference service. Check the error details.

Failed to call the inference service. Check the error details.

IIT.0201

The input param is invalid!/The input param is invalid, please check your key!

The request parameter is invalid.

Check whether the request parameters are correct.

IIT.0202

Interval Server Error!

An internal error occurred.

Contact technical support for assistance.

IIT.0203

The input param is invalid, the input data lens is less than the train data lens!

The request parameter is invalid. The data length in the input parameter is less than that used for training.

Check that the feature name and number of features in the request body are the same as those in the training data.

PREDICT.0102

"Json format is wrong!" or other data-related errors.

The request data is not in JSON format.

Other data-related errors occur.

Set the request body to be in JSON format.

Adjust the request body based on information about the data-related errors.

PREDICT.0201

The input param is invalid!/The input param is invalid, please check your key!

The request parameter is invalid.

Check whether the request parameters are correct.

PREDICT.0202

Interval Server Error!

An internal error occurred.

Contact technical support for assistance.

PREDICT.0203

The input param is invalid, the input data lens is less than the train data lens!

The request parameter is invalid. The data length in the input parameter is less than that used for training.

Check that the feature name and number of features in the request body are the same as those in the training data.

APIG.0101

The API does not exist or has not been published in the environment.

The API does not exist or has not been published.

  • Check whether the API URL is correct. For example, check whether the project ID is included in the URL.
  • Check whether the HTTP request method (such as POST or GET) is correct.

APIG.0201

Backend timeout.

Request timed out.

  • Check whether the API call requests are initiated too frequently. If so, check the return value in the code and resend the requests later (for example, 2 to 5 seconds later). You can also check in the backend whether the result of the previous request is returned. After the result of the previous request is returned, send the next request.
  • Confirm with technical support to check whether the API has been deployed.

APIG.0301

Incorrect IAM authentication information.

The IAM authentication information is incorrect.

  • decrypt token fail: The token fails to be parsed.
  • token expires: The token has expired.
  • verify aksk signature fail: The AK/SK authentication fails.
  • x-auth-token not found: The x-auth-token parameter is not found.
  • If the token fails to be parsed, check the method for obtaining the token, whether the request body is correct, whether the token is correct, and whether the environment for obtaining the token is the same as the environment for calling the API.
  • If the token has expired, obtain a new token that is valid permanently.
  • Check whether the AK/SK pair is correct. For example, check whether the SK for the AK is correct and whether an extra space is included in the AK/SK pair.
  • AK/SK-based authentication errors occur frequently. If an AK/SK pair fails to be authenticated for more than five consecutive times, the AK/SK pair is locked for 5 minutes (the AK/SK-based authentication is considered as an abnormal authentication request within 5 minutes). After 5 minutes, the AK/SK pair is unlocked and re-authenticated.
  • Check that the spelling of the value for X-Auth-Token in the request header for an API call is correct.

APIG.0308

The throttling threshold has been reached: policy user over ratelimit,limit:XX,time:1 minute.

The request exceeds the default rate limit of the service.

  • Use the retry mechanism to rectify the fault by checking the return value in the code and retrying the requests after a short period of time (for example, 2 to 5 seconds).
  • Check in the backend whether the result of the previous request has been returned. If it has, send the next request. This helps prevent excessively frequent requests.