This depends highly on what you put onto the API. Let’s say you ask 03-mini to greet you with a short “hi” every time you type “1”.
I mean the API behind it is paid based on how many tokens each request needs and I would assume that ChatGPT has to pay money to use the API. Which then means when you push a novel on each request you will produce higher costs for them (and also use up a swimming pool of water per day for cooling and energy doesn’t grow on trees either).
See also how OpenAI is structured:
https://openai.com/our-structure/
Just to put things in perspective:
I am using o3-mini over the API when I run out of plus requests and just for personal chatting with it I pay up to 15$ per day (should get pro maybe).
I think when they put something like “you have X token left until you get into pause mode” would still be nice. But that is also not easily to program if you ask me… and it would produce more costs for servers that keep track on that for hundrets of millions of users - which then might lead to even less requests - so you could trade requests for transparancy - I don’t think anyone really would want that.
And I would also assume that they give you some extra token when servers are not fully loaded. With transparency this would not be possible. So I am absolutely against it and trust them that what you get when you run into limits constantly is a loss for OpenAI or ChatGPT or whoever has to pay what to avoid taxes and regulation problems of the Nonprofit structure 