Unknown timeout using Lambda Function along with API Gateway to await a response from Openai for longer than 1 minute

I have requested an account-level timeout limit for API Gateway to extend to 2 minutes and have enforced the 2-minute timeout limit on it. I have also ensured that the Lambda Function does not time out until 10 minutes later. The following is some proof:

I have also enabled logging on the API Gateway level and everything works fine and there's no errors being logged:

No issues are spotted on the Lambda Function either:

In short, no errors can be logged or observed while a response is indeed returned from Openai to the Lambda Function and the API Gateawy, where responses can be gotten. The openai response usually takes 30 seconds to 1 minute. What is the reason that the frontend logs an 503 Error Status while the backend logs no errors? When I run the backend in the local, since the response can be awaited forever, no errors are logged. Given the sufficient timeout limits I have set for the two resources, what is the cause of this error?

Many thanks for any helpful advice.