[Tixx7]

I switched to 2.5 Flash (non-think) for most of my projects because it was such a good model with good pricing.

Cost is an important factor so hoping that flash-lite is sufficient, even tho its somtimes more than 50% worse in relevant benchmarks which sucks.

Was also just looking at 4.1-mini but thats more expensive and often scores around the same as flash-lite in benchmarks (except coding which i dont care about).

Crazy to think that even after this move by google, openai is still the worse option for me, at least regarding API. Other than API im actually using chatgpt (o3/o4-mini, 4o is a joke) a lot more again lately after 2.5 Pro got nerfed