I’ll be honest — I’m done with Gemini models.
I don’t understand how a company with that much user data and nearly the entire surface web at its disposal still can’t ship a genuinely intelligent model.
I’m a Google Pro annual subscriber, mainly for the AI Studio credits and quota. First they killed the credits. Then they gutted the quota for the only models worth using — Claude Sonnet 4.6 and Claude Opus 4.6. Those were the two reasons I stayed.
But my bigger issue is: why is Google even making these models? If you only care about speed, build that for your search engine. Some of us want intelligence. Having Gemini in my IDE and watching it fail at basic tasks is painful. I don’t fully vibe-code, thank god — with 5 years of programming experience, I can catch its mistakes. Otherwise I’d be suffering.
Honestly, Google should either get serious or step out of AI. It feels like they’re just releasing models to stay in the race. Meanwhile they’re scraping data from small creators, paying giants like Reddit for theirs, and still shipping underwhelming products.
And Gemini 3.6 Flash? Benchmarks look fine, but real-world capability tells a different story. It’s just a slightly more efficient version of 3.5 Flash, which was already underwhelming for the price. At this point, Google could learn a thing or two from Chinese open-source models.