Persistent Gemini API Location Error — HTTP 400 FAILED_PRECONDITION

Gemini API
1

Subject: Persistent Gemini API regional access error despite successful connectivity and supported-region IP

Hello Google AI Support Team,

Thank you for your response.

I have performed additional diagnostic tests, but the issue remains unresolved.

I am developing an AI-powered legal assistant using the Gemini API, Python, and Supabase. I use the gemini-embedding-2 model to generate embeddings for a legal document search system.

The API consistently returns the following error:

HTTP 400 — FAILED_PRECONDITION

User location is not supported for the API use.

I have now confirmed the following:

  1. My Python script can successfully connect to Google’s API endpoint.
  2. My external IPv4 address is geolocated to Ashburn, Virginia, United States, through the FDCserversnet network.
  3. The same error occurs when I query the general Gemini API model list endpoint (/v1beta/models), without requesting embeddings.
  4. The error persists when calling the API using gemini-embedding-2:embedContent.
  5. I previously tested VPN endpoints in Finland and the United States, but the error remained unchanged.
  6. The API key is successfully read from my local environment file. I have not included the key in this report.

Since the error occurs both when listing models and when generating embeddings, it does not appear to be limited to the embedding model itself.

Could you please investigate whether this error may be caused by an account-level or Google Cloud project restriction, an incorrect regional eligibility determination, or another server-side access control?

If possible, please clarify which eligibility check is failing and what specific action is required to restore API access.

I would appreciate a technical investigation rather than further general recommendations to change VPN servers, as I have already tested multiple locations.

Thank you for your assistance.

Best regards