The API documentation at OpenAI is currently one behemouth HTML document: https://platform.openai.com/docs/api-reference.
One of the ironies of that is that it means you can’t easily feed it into an LLM for assistance. Normally I would @ cursor into a page of documentation but because the page is so massive it’s too large for its context window. So instead I have to cut and paste individual sections out. Not only is that slow but the sheer size of the page means it slows even the browser down during navigation.
Please can you split the documentation into pages that reflect smaller subsections of it. These will be quicker to load and can easily be ingested into Cursor (or ChatGPT).
One area that’s particularly tough is the Realtime API. There are so many events and the documentation is so thin that it requires a lot of research to understand and I find myself creating duplicate documentation in Cursor just so as there’s a way to navigate it more clearly.
Thank you