As an Ultra user, I really can’t take this anymore. Last month, the Opus quota was plenty—I could write code for 3-4 hours and still have 80% left. But this month, the quota is extremely low and runs out so easily. Just now, I kept track: after making fewer than 20 Opus requests, my quota is already down to 20%. I want to ask, is Gemini trying to abandon us users? I understand you want to lower quotas, but less than 30 Opus requests every 6 hours—are you serious? If this is how it’s going to be, you should just shut down your ridiculous Ultra plan as soon as possible. From a disappointed user. I think I’m switching to Claude Code. Goodbye.
5 Likes
i will quit soon，the ultra is a joke
3 Likes
Hold the door open, I am on my way out too!
This was the last straw for me:
2 Likes
I’m also ready to close the Ultra Plan
1 Like
I am out too with my Ultra plan! Got to a point where I can not anymore tolerate it
I’ve downgraded. and had to completely remove all my MCPs, reduce skills, and completely rewire my systems to handle the reduced bad usage. It’s unacceptable relative to peers so I’ve gone back to other IDEs.
Antigravity was never that good compared to Cursor or even VS Code with Copilot. The only thing it had over them was extremely generous usage limits. The tool use was horrible, it ate an insane amount of RAM compared to baseline VS Code, and was always so buggy. Google is very far behind on agents and integrating its otherwise wonderful models into any product, yes. But Antigravity and Gemini CLI are the bottom of coding tools. I just got Copilot instead
3 Likes
I just switched to Claude Code $200 Per Month subscription and then they also decreased the limits lol
Perhaps the “running at a cost to grab market share” strategy hasn’t worked out as well as everyone hoped.
how CC Max limits compare to AG Ultra now?
Yeah seems that they are pulling the rug on us. It seems so deceptive to bring in users willing to pay for the highest tier plan – letting them experience the latest models with what felt like great limits. Then after having them pay for 6 months of the service - completely pull the rug and not giving them nearly what they once did. And the lack of visibility into all of this – how do they determine usage? Tokenizer? Units of Work? Like how do we know what we are getting and how can we measure consistency.
it’s definitely higher for Claude Opus 4, in Claude Code. since they have lowered it, I have not been impacted at all. So I think the impact, has, gone to all the Pro subscribers or maybe the Max 5X subscribers.
1 Like
thanks, will try soon