It’s absolutely absurd how my Opus 5.5 limits got drained in 10 minutes, for a simple redesign process
This is unacceptable; the Gemini models are horrible. I expect nobody to reply to this from Google, but why bother adding the new models if you’re going to make it unusable on an almost 20 euro plan?
Facing the same issue. 91% quota got burned in 28 minutes for less than 7 file reads and edits. Gemini 3.6 to 3.8 are crazily bad these days.
Check replies on one of my reports on the same issue you mentioned, looks like bots 
There are two “issues” in reality, but they are intended behavior:
If you really need nice quota with Opus 5.5 then get a plan with the competitor.
Cheers.
Hello everyone.
I have been using this platform—specifically the “PRO Plan”—for just under a month, and it has been working excellently for me; I rely primarily on the Gemini models.
Here is my candid opinion based on practical experience:
This platform outperforms the latest global platforms. However, it requires expert guidance from you—as the developer—to steer these highly intelligent, professional models toward achieving extreme accuracy and delivering substantial results.
Let me explain how I develop and work within this high-quality environment.
To begin with, I have tried various software tools, but I only found a truly comfortable and efficient workflow on this platform. That said, based on my experience and how I have adapted the platform to deliver real outputs in record time—with only minor, easily fixable errors—there are important steps you should follow.
First, create a daily development plan. Divide your total weekly quota across the seven days (averaging about 14% per day). In fact, the platform handles this task for you;. The counter located below the Gemini quota section tracks usage rates, so you do not even need to do the calculations yourself. (Since I am developing new software, the process requires more effort and experimentation than usual to reach the final product.)
For my needs, this daily limit is sufficient, and I use the “medium” interaction setting with the model. Regarding Google subscription plans, I recommend upgrading to the highest Google One tier to leverage maximum structural capabilities and ensure unrestricted processing depth.
Secondly, you need to rein in these “digital beasts.” They possess neither feelings nor fear, nor do they realize when your quota might run out. You need to input a custom prompt—tailored specifically to your project and software—that is rigorous and precisely directprecise. Then, observe the results.
Thirdly, regarding models other than Gemini—such as Sonnet and Opus:
Do not use these models for routine coding tasks. Instead, employ them when dealing with critical or pivotal aspects of your application; they are designed for deep reasoning rather than standard coding or everyday development work. By following these methods, you can develop a high-performance application, whether for the web, Android, or any other target platform. I am available to answer any questions you may have, so please do not hesitate to reach out. Thank you very much, and apologies for the length of this message.
Best regards,
Sami
Hey, . I have the 200 dollar ultra plan and they reduced it too..
They haven´t reduced anything, the quota is the same as before (about 1-3 high opus prompts), as intended. Don´t misread me please.
Cheers.
If you want to use Opus or Sonnet, use the Anthropic subscription. Gemini models give decent quota, and you can use them all day long without issues. 3.8 flash is also quite powerful, so use it for your tasks and you’ll get a lot done
No, no it is not. Weekly limits used to be 4x (25% → 25% → 25% → 25%), meaning using a full five hour charge would take up 25%. Now, it’s 2x (50% → 50%) and is much easier to actually exhaust it (24 hours for me as opposed to 4 - 6 days prior). No need to lie.