Hello,
We are developing an educational app for children aged 7–10 (iOS/Android) and would like clarification on the “Age Requirements” clause of the Gemini API Additional Terms of Service (effective March 23, 2026).
How we use the Gemini API
- Only on the developer side, offline, in batch, to create static assets: (1) audio clips from fixed scripts (gemini-3.8-flash-tts), (2) character and scenario images (Gemini image generation), (3) test data / example answers (a Gemini text model).
- The generated files are bundled inside the app as static files.
- The app itself never calls any Gemini API at runtime.
- No end-user data (and no data from children) is ever sent to the API. Our prompts contain only our own scripts and image descriptions.
- Our project is on the paid tier with billing enabled.
The clause says we “will not use the Services as part of a website, application, or other service … that is directed towards or is likely to be accessed by individuals under the age of 18.”
Questions
- Is it permitted to include audio and images generated this way (offline, developer-side) in a children’s app, given that the app itself does not use the Services at runtime?
- If not, which product or terms would allow generating such static assets for a child-directed app (for example, Cloud Text-to-Speech under Google Cloud terms, or Vertex AI)?
- Are there attribution or other requirements for generated audio and images bundled in an app?
- Do the same terms apply to image generation and text generation used in the same way?
Related question from another developer: Age Restrictions in Gemini API Terms: general-audience app (rated 3+) with an adult-only AI feature
We do not plan to call the API at runtime. An official answer would help other developers in the same situation. Thank you!