There’s a lot of articles about it. It costs them $500+ for heavy users. They do this to capture market share and also to train their agent loops with human reinforcement learning.
You chose a bad one. It just asserts the 95% figure without evidence and then uses it as the premise for the rest of the article. That just confirms what I said earlier about how "Every time I see claims on profitability it's always hand wavy justifications.". Moreover the article reeks of LLM-isms.
Source? Every time I see claims on profitability it's always hand wavy justifications.