Fail to load traces when using a different API key for tracing

[LAHARI2003]

https://openai.github.io/openai-agents-python/config/#:~:text=You%20can%20specifically%20set%20the%20API%20key%20used%20for%20tracing%20by%20using%20the%20set_tracing_export_api_key%20function.

for doing the api calls i have used the export (api key 1) and set my trace using another api key say (api key 2),
trace is seen but there is an error being throwed every time i try to check the POST/v1/responses of an agent

[LAHARI2003]

The trace was visible when we used Chat Completion end point through

export OPENAI_API_KEY  =  ( api key 2 )
openapi_client = AsyncOpenAI(
    api_key="( api key 1 )",
    base_url="https://api.openai.com/v1"
)
 model_1 = OpenAIChatCompletionsModel(
     model = "gpt-4.1-mini" ,
     openai_client = openapi_client
)

this case the trace was visible as expected in the dash board of api key 2

But the above did not work when used with the Responses one by that i mean

export OPENAI_API_KEY  =  ( api key 2 )
openapi_client = AsyncOpenAI(
    api_key="( api key 1 )",
    base_url="https://api.openai.com/v1"
)
 model_1 = OpenAIResponsesModel(
     model = "gpt-4.1-mini" ,
     openai_client = openapi_client
)

the issue persisted in the this case which shouldnt have happened

@rm-openai @seratch
where do you suggest the issue is ?

[imakshat47]

It appears that the issue arises because Trace retrieves Response API logs from the OpenAI platform using the response ID. When non-OpenAI models are used, the corresponding response data is not available on the OpenAI platform, resulting in a 404 error.

Thanks @LAHARI2003 for direction of finding response id.