Hello,
I'm running a production application on Vertex AI that calls `gemini-3.1-flash-image` via `generateContent` (image generation) using the `global` endpoint, and I'm consistently hitting `429 RESOURCE_EXHAUSTED` after just a handful of requests.
I've thoroughly checked the Quotas & System Limits console for my project and confirmed:
- The only quota row directly tied to this model ("Generate content requests with image input per minute per project per region per base model and resolution for global endpoint") is marked "Adjustable: No" and shows 0% usage even while I'm receiving 429s.
- The generic "Global generate content requests per minute per project per base model" quota list (which shows adjustable per-model limits for dozens of other models, including several experimental ones) does NOT include a row for `gemini-3.1-flash-image` at all.
This suggests the model's rate limit isn't yet exposed through the standard self-service quota system.
Could someone from the Vertex AI team help allowlist / raise the per-minute request quota for this model on my project?
Project ID: alquimia-b19..
Project number: 980792940..
Model: gemini-3.1-flash-image
Endpoint/region: global
Use case: AI-generated digital art product for end customers (Brazil-based small business). Each customer purchase triggers a batch of ~10 image generations. Expecting on the order of tens of paying customers per day at launch, scaling from there — so I need a request-per-minute limit well above the current default (which appears to allow only a very small handful of requests before hard-failing) to serve customers reliably without long waits between generations.
Happy to provide any additional information needed. Thank you!