Errors

When calling the API, the response code consists of two parts: the outer layer is the HTTP status code, and the inner layer is the business error code defined by Z.AI in the response body, which provides a more detailed error description.

HTTP Status Code

Code	Reason	Solution
200	Business processing successful	-
400	Parameter error	Check if the interface parameters are correct
400	File content anomaly	Check if the jsonl file content meets the requirements
401	Authentication failure or Token timeout	Confirm if the API KEY and authentication token are correctly generated
404	Fine-tuning feature not available	Contact customer service to activate this feature
404	Fine-tuning task does not exist	Ensure the fine-tuning task ID is correct
429	Interface request concurrency exceeded	Adjust the request frequency or contact business to increase concurrency
429	File upload frequency too fast	Wait briefly and then request again
429	Account balance exhausted	Recharge the account to ensure sufficient balance
429	Account anomaly	Account has violation, please contact platform customer service to unlock
429	Terminal account anomaly	Terminal user has violation, account has been locked
434	No API permission, fine-tuning API and file management API are in beta phase, we will open soon	Wait for the interface to be officially open or contact platform customer service to apply for beta
435	File size exceeds 100MB	Use a jsonl file smaller than 100MB or upload in batches
500	Server error occurred while processing the request	Try again later or contact customer service

Business Error Codes

Error Category	Code	Error Message
Basic Error	500	Internal Error
Authentication Error	1000	Authentication Failed
	1001	Authentication parameter not received in Header, unable to authenticate
	1002	Invalid Authentication Token, please confirm the correct transmission of the Authentication Token
	1003	Authentication Token expired, please regenerate/obtain
	1004	Authentication failed with the provided Authentication Token
	1100	Account Read/Write
Account Error	1110	Your account is currently inactive. Please check your account information
	1111	Your account does not exist
	1112	Your account has been locked, please contact customer service to unlock
	1113	Your account is in arrears, please recharge and try again
	1120	Unable to successfully access your account, please try again later
	1121	Account has irregular activities and has been locked
API Call Error	1200	API Call Error
	1210	Incorrect API call parameters, please check the documentation
	1211	Model does not exist, please check the model code
	1212	Current model does not support ${method} call method
	1213	${field} parameter not received properly
	1214	Invalid ${field} parameter. Please check the documentation
	1215	${field1} and ${field2} cannot be set simultaneously, please check the documentation
	1220	You do not have permission to access ${API_name}
	1221	API ${API_name} has been taken offline
	1222	API ${API_name} does not exist
	1230	API call process error
	1231	You already have a request: ${request_id}
	1234	Network error, error id: ${error_id}, please contact customer service
	1261	Prompt exceeds max length
API Policy Block Error	1300	API call blocked by policy
	1301	System detected potentially unsafe or sensitive content in input or generation. Please avoid using prompts that may generate sensitive content. Thank you for your cooperation.
	1302	High concurrency usage of this API, please reduce concurrency or contact customer service to increase limits
	1303	High frequency usage of this API, please reduce frequency or contact customer service to increase limits
	1304	Daily call limit for this API reached. For more requests, please contact customer service to purchase
	1305	The API has triggered a rate limit.
	1308	Usage limit reached for . Your limit will reset at ${next_flush_time}
	1309	Your GLM Coding Plan package has expired and is temporarily unavailable. You can resume using it after renewing the subscription on the official website. https://z.ai/subscribe
	1310	Weekly/Monthly Limit Exhausted. Your limit will reset at ${next_flush_time}
	1311	Your current subscription plan does not yet include access to ${model_name}
	1312	This model is currently experiencing high traffic. Please try again later, or switch to another model such as ${model_name}
	1313	Your usage violates the Fair Use Policy. Your request rate has been restricted. See Subscription Service Agreement for details. To restore access, go to Personal Center → My Subscription and request to lift the restriction

Error Shapes

Errors are always returned as JSON, with a top-level error object that includes a code and message value.

{
  "error": {
    "code": "1214",
    "message": "Input cannot be empty"
  }
}

Error Example

The following is the response message of a curl request, where 401 is the HTTP status code and 1002 is the business error code.

* We are completely uploaded and fine
< HTTP/2 401
< date: Wed, 20 Mar 2024 03:06:05 GMT
< content-type: application/json
< set-cookie: acw_tc=76b20****a0e42;path=/;HttpOnly;Max-Age=1800
< server: nginx/1.21.6
< vary: Origin
< vary: Access-Control-Request-Method
< vary: Access-Control-Request-Headers
<
* Connection #0 to host open.z.ai left intact
{"error":{"code":"1002","message":"Authorization Token is invalid, please ensure that the Authorization Token is correctly provided."}}

Note: When using streaming (SSE) calls, if the API terminates abnormally during inference, the above error codes will not be returned. Instead, the reason for the exception will be provided in the finish_reason parameter of the response body. For details, please refer to the description of the finish_reason parameter.