Dear Google AI Support Team,
I am writing to report a persistent technical issue regarding Rate Limit (429 Resource Exhausted) errors I am encountering with the Gemini 2.5 Flash model.
Despite my account being confirmed as Tier 3 (Paid) and my total daily usage being well under 5% of the allocated quota, I am frequently blocked by 429 errors. I have confirmed that each of my projects uses a dedicated API key to avoid cross-project rate limiting.
System Details:
-
Model: Gemini 2.5 Flash
-
Account Tier: Tier 3 (Paid)
-
Project ID: vsf-vapp1-prod-12
-
Environment:
Python SDK
Problem Description: The errors occur even during periods of moderate request frequency. It appears the system might be triggering a “Resource Exhausted” flag prematurely, or there is a discrepancy between the Tier 3 limits and the actual enforcement on my Project ID.
Supporting Evidence:
1. Quota Dashboard Overview: This image shows that my overall daily usage is significantly below the limit.
2. Peak Usage & Error Spikes: The 429 error spikes in the metrics dashboard during the time of the incident.
3. Billing/Tier Confirmation: Confirmation that my account is correctly recognized as Tier 3.
I would appreciate it if you could investigate whether there is a backend synchronization issue with my project’s rate limits or if my requests are hitting a specific regional capacity constraint.
Looking forward to your prompt response.
Best regards,
HieuBQ