In AI Studio, it plowed through stuff more intelligently than 3.1 Pro, way faster, and I didn't get rate limited which normally would have happened at least three times due to how many tokens I spent yesterday.
Crazy how many people are pooh-poohing Gemini while I'm over here "feeling the AGI" and "tokens too cheap to meter". (yeah I know it is subsidized)
What are your main drivers Coeurnix? And what engine are you working in?
Excellent! I’m glad it’s working so well for you! I know there are a lot of cynics about AI being too expensive soon (and, if Deepseek hadn’t happened, maybe history would have gone that route for awhile), but I think it’ll go the way of Internet, phone call, and text usage limits: more and cheaper as time goes by.
I mainly use Codex CLI, but have a Z.ai sub that I use with Claude Code for alternative ideas or if I’ve used my week’s worth of GPT. I’m using three.js for my game right now, though I have a second idea that will have a simple custom engine if there’s time. Thanks for asking!