AI error code parsing
When Verity fails to call the LLM service, the corresponding HTTP error code is returned. This article lists common error codes and what they mean to help you quickly locate the problem.
Client error (4xx)
The request parameters sent to the API are incorrect, such as malformed JSON, missing required fields, or invalid parameter values. Often caused by client configuration issues.
The API Key is invalid, expired, or not provided. Please check that your key is correct or regenerate it in the service provider console.
Key is valid but does not have permission to access the requested resource. It may be that the account has not completed real-name authentication, or the selected model is not activated.
The requested model or interface address does not exist. Please check that the Base URL and Model name are correct.
The server did not receive the complete request within the specified time. It may be due to network instability, it is recommended to check the network and try again.
The request is in the correct format, but the content is not valid, such as the message content is too long, contains unsupported parameters, etc.
Too many requests were sent in a short period of time, triggering the service provider's rate limit. We recommend waiting 1–2 minutes before retrying, or switching to another AI provider.
Server-side error (5xx)
An exception has occurred within the service provider. This is not your issue and is usually fixed by the service provider in a short period of time. Please try again later.
The gateway received an invalid response from the upstream server. It is usually a temporary failure of the service provider, you can try again in a moment.
The service is temporarily overloaded or under maintenance. Please try again later. If it persists, try switching to a different AI provider.
The gateway timed out waiting for the upstream server to respond. It may be that the service provider is loaded, it is recommended to wait and try again.
Troubleshooting suggest
- encountered429When, wait a minute or two and try again, or switch directly to another AI provider.
- encountered503, you usually only need to try again to recover.
- Prioritize checking that the API Key and Base URL are configured correctly when a 4xx error persists.
- If all providers are unavailable, it may be a local network issue, check your network connection or proxy settings.