Hi Google AI Developers Team,
I am a student trying to build and launch a startup, and I am currently paying for a premium Google AI Pro subscription. I am writing to report a highly frustrating user experience and model limitation within the Google Antigravity ecosystem.
Issues Faced:
-
Missing Reasoning Models: The advanced Gemini Pro models (like Gemini 3.1 Pro) have completely vanished from my selection dropdown. Furthermore, Gemini 4 Argon is entirely missing on the Pro tier for the terminal workspace.
-
Locked into Basic Flash: As a paid subscriber, I am heavily restricted to using Gemini 3.8 Flash. This model is simply not capable of handling complex logic, code generation, and deep debugging required for building software architectures.
-
Unusable Third-Party Options: While third-party models like Claude Sonnet/Opus 5.5 and GPT-OSS 120B are visible, they are capped at “Medium” credit thresholds. Due to standard code files having large token footprints, these models exhaust their allowance and cut me off after just a single prompt.
As a student developer investing in your Pro tier, being left with a single baseline Flash model makes it impossible to advance my work. I urgently request that the engineering team reviews the model allocation policy for Pro subscribers so that advanced reasoning models are properly provisioned for intensive terminal workloads.
Thank you.