It is sufficient to say that the output of gpt-5, even aligned as a product support specialist, was not great when sending this message…and not worth repeating. Find a better AI to get your answer with this input.
You’re going to help me answer this developer’s question:
(your post)
The documentation they didn’t find:
GPT‑5 in the API platform is the reasoning model that powers maximum performance in ChatGPT. Notably, GPT‑5 with minimal reasoning is a different model than the non-reasoning model in ChatGPT, and is better tuned for developers. The non-reasoning model used in ChatGPT is available as gpt-5-chat-latest.
My additional notes
This model is offered for experiments, or a different way to do some benchmarks, but it is tuned to be “chatty” and write markdown into ChatGPT’s renderer, instead of plain text requiring the developer specifically to ask for markdown. It is like the previous chatgpt-4o-latest - the current model used by ChatGPT, but not really for developer use. Not only does the chat model not reason internally, but also, tool calling and structured outputs are disabled, kind of discouraging development.
Response
Produce a friendly forum answer in my voice.