Hello Google AI Team and fellow developers,
I am raising a massive red flag regarding the current state of the Gemini API infrastructure, specifically concerning the Nano Banana models and their integration with third-party partners.
-
The Bottleneck: I am a paying Adobe subscriber utilizing the newly integrated Google Nano Banana 2 partner model inside the Adobe Firefly interface. For the past week, almost every attempt to generate an image—which costs 80 generative credits per prompt—results in an immediate failure: “Something went wrong. This model is experiencing heavy load.”
-
The Unacceptable Cause: It is painfully obvious that the issuance of excessive free trials and unrestricted free image generations on Google’s native platforms is actively cannibalizing the server infrastructure. The servers are collapsing under the weight of free-tier traffic, and it is entirely crippling the API pipeline for enterprise partners.
-
The Financial Impact: As a user spending actual money and credits through an integrated platform, my resources are effectively being burned. If Adobe is paying Google for API access, and I am paying Adobe for credits, why is my API request being throttled by the same server load as a free user? There is zero priority routing.
-
The Ultimatum: Google needs to make a definitive choice to protect its API partners. You must either aggressively expand your data center capacity to handle the actual demand of these models, or immediately throttle/stop the free generations. Paid API requests routing from enterprise partners MUST be prioritized.
We need scalable infrastructure and proper load balancing, not a perpetual traffic jam that wastes paid credits. Please address this backend routing issue immediately.
