“What you need to do
Manually verify and review your current usage to plan ahead and prevent service disruption when the new caps take effect:”
Service disruption? Caps?
Why can’t google cloud / ai just charge us and let us pay?
This “Gemini API usage tier updates and billing caps”, makes no sense.
What’s the use case?
What’s the reasoning?
How does this help developing on Gemini?
Recently I had to create a support ticket with Google Cloud in order get permission to add projects to our billing account. Adding projects = revenue for Google Cloud but only if they agree to let us pay them. Insane.
Best Regards,
Donald
We don’t use AI studio at all, does this billing cap still apply to API usage through normal GCP Gemini APIs also?
I’ve been researching this problem because I’m building a small tool around it.
What I keep seeing is that spend controls are improving provider by provider, but for small teams the problem is still messy:
- different providers
- different billing behavior
- different key exposure risks
- and not always a simple “hard stop” mental model
I’m building HardCap as a simple spend firewall for AI/API keys:
Main idea: define a hard limit per key/project and block requests when it’s reached.
If anyone here has dealt with leaked keys, runaway usage, or delayed alerts, I’d love feedback.