I am developing a travel translation web application called Fortress Translator.
The application is operated from Japan. The architecture is:
User in Hong Kong → my web server in Japan → Gemini API
The Gemini API key is kept only on my server in Japan. I am the API customer and I pay the Gemini API usage fees. End users do not receive a Google account, Gemini API key, or direct API access. They only submit text to my translation application and receive translated text.
I would like an authoritative clarification about the Gemini API regional restrictions.
- Can a Japanese resident temporarily traveling in Hong Kong use this application while physically located in Hong Kong?
- Can a Hong Kong resident use the same application?
- Do the Gemini API regional availability restrictions apply to the physical location of the end user of an API-powered application, even when all Gemini API requests are sent from a server located in Japan?
- If either case is not permitted, is there an exception, approval process, or supported architecture that would allow this type of service to be offered to users in Hong Kong?
I am not asking how to bypass geographic restrictions, and I am not asking whether the API request technically succeeds.
I need clarification about whether this specific architecture is permitted under the current Gemini API terms and regional policy.
If this forum cannot provide an authoritative answer, please direct me to the appropriate Google team or official support channel that can provide written confirmation.