I am experiencing a persistent “Internal Error” in Google AI Studio as soon as I switch to my Paid Tier (Tier 1) API key. In the Free Tier, everything works perfectly. When using the Paid Key, stable models (like 1.5 Flash) occasionally work with high latency, while others (like Gemini 3 Flash Preview) fail 100% of the time.
What I have already tried/verified:
-
Project Setup: Created a completely new Google Cloud Project to avoid “auto-generated” project issues.
-
Billing: The project is actively linked to a verified billing account. No budget restrictions are blocking the usage (confirmed via GCP console).
-
API Activation: The “Generative Language API” has been deactivated and reactivated to refresh the billing status.
-
API Keys: Created fresh API keys directly within the new GCP project and imported them into AI Studio.
-
Safety Settings: All safety filters are set to “Block none”.
-
Quotas: GCP Console shows “Unlimited” for Tier 1, but logs show a high error rate (up to 66%) or simply “No data found” despite failed requests in AI Studio.
Observation: The error seems to be related to the backend validation of the Paid Tier. It feels like my account is “stuck” between tiers or being throttled by a silent security check.
Project ID: Region: Europe (Germany)
Could you please check if there is a manual verification hold or a synchronization issue on the backend for my Tier 1 access?