I’m literally just trying to use openai I don’t even have a server, could you please try testing a gpt model from open so using one of the models busy just doing it through custom LLM route and using the open so chat completion endpoint? That’s all I need