Maybe it’s just me, but Gemini 3 Flash and Gemini 3.1 Pro (High) are so bad after your last update. Gemini 3 Flash is just horrible now. I can’t fix even a small bug anymore. Basically it’s useless now. Gemini 3.1 Pro (High) is just a little bit better.
Could it be that someone was recently hired who might be intentionally causing issues for the company?
Because it’s just so obvious.
Recent changes are just so negative.
Peopel can’t work with the system anymore.
Don’t you see that?
I concur with you on this. Gemini has created more trouble than helping me to do things. even simple instructions to revert changes in the current working branch led to reverting changes in the entire workspace, without even confirming once with the user. DOES NOT FOLLOW the rules in the GEMINI.md at all. for simple dockerfile fixes, it starts fixing code - totally unrelated to the errors observed.
It was never the better than Claude Sonnet 4.6 but at least it was usable before, now it’s become unusable for me.
So true, same
3.1 Pro already had a problem with it constantly trying to fix and redo things it had already worked on, and now it’s become five times more frequent. I don’t use the API for programming (Haha, I conduct research on various topics + rp in my free time), but it’s driving me crazy, in a bad way.
Previously, I could at least see differences from the 3 flash, but now they are literally the same, although the cost is strikingly different. It feels like they’re trying to sell us cheaper models under the guise of “3.1 Pro.”
At the start the situation was much better. Maybe this is an optimization problem on their part, I don’t know how else to explain it?..
they pushed me to unsubscribe and start learning other models and tools, , even thoug minimax 2.5 is not as advanced, its more reliabel, i use it for cheap now
