fix: use gpt-4o-mini for validating credentials (langgenius#9387)

api/core/model_runtime/model_providers/openai/openai.py

@@ -19,9 +19,9 @@ def validate_provider_credentials(self, credentials: Mapping) -> None:
         try:
             model_instance = self.get_model_instance(ModelType.LLM)
 
-            # Use `gpt-3.5-turbo` model for validate,
+            # Use `gpt-4o-mini` model for validate,
             # no matter what model you pass in, text completion model or chat model
-            model_instance.validate_credentials(model="gpt-3.5-turbo", credentials=credentials)
+            model_instance.validate_credentials(model="gpt-4o-mini", credentials=credentials)
         except CredentialsValidateFailedError as ex:
             raise ex
         except Exception as ex: