Comment by caminanteblanco

Ok, one issue I have with this analysis is the breakdown between input and output tokens. I'm the kind of person who spend most of my chat asking questions, so I might only use 20ish input tokens per prompt, where Gemini is having to put out several hundred, which would seem to affect the economics quite a bit