Hi,
I’m a Google AI Pro subscriber and I’ve noticed a serious inconsistency in model quota refresh times within Antigravity.
As shown in my screenshot:
- Gemini 3.1 Pro (High) → Refreshes in 5 days, 4 hours
- Gemini 3.1 Pro (Low) → Refreshes in 5 days, 4 hours
- Gemini 3 Flash → Refreshes in 4 hours, 59 minutes
- Claude Sonnet 4.6 (Thinking) → Refreshes in 4 hours, 59 minutes
- Claude Opus 4.6 (Thinking) → Refreshes in 4 hours, 59 minutes
- GPT-OSS 120B (Medium) → Refreshes in 4 hours, 59 minutes
All other models refresh within ~5 hours as expected, but Gemini 3.1 Pro (both High and Low) is locked out for over 5 days. This makes it practically unusable for a paying subscriber.
