You are a gentleman! There doesn't appear to be any documentation in the llama-index docs yet but there is support added via the langchain module. It looks like I can "from langchain.llms import OpenAIChat" and then use this function to build a new index using "gpt-3.5-turbo" model. I will give this a go and see if it works. I will look into Treeindex too, reading the docs around these different indexing tools was getting a bit too complex for me