It’s quite UGLY that the Gemini-3.5-FLASH models share a quota with GEMINI-3.1

It’s unfair that in Antigravity CLI, the Gemini-3.5/FLASH models share a quota with the lower-end GEMINI-3.1/FLASH/PRO models, while in GEMINI CLI, when you use the 3.1PRO model and your quota for that model runs out, you can chose to 3.1FLASH or even an even lower-end 3.0FLASH/PRO model.

This should be changed; it’s very unfair.

I used it today; two queries consumed 10% of the quota, so it feels like the limit has been slashed again. Based on this calculation, the quota allows for only about 20 queries over five hours. I was using the “Flash Low” model. I hope Google isn’t quietly cutting the limits.