Eigen RadarAI
Analysis

Gemini API turns on request storage for new paid projects

Google changed the request-storage default for new paid Gemini API projects on October 26. Generate Content calls are now stored unless developers choose otherwise, while existing projects keep their previous settings. A request-level switch or the project’s logging controls can turn storage off. The change affects how application developers handle request data when creating a new project.

Artificial Intelligence··Evening
In an imagined recording machine, a metal drum touches cream tiles beside a clear cassette holding matching tiles and an upright control lever.

New paid projects begin with storage enabled

Google changed the default for new projects on the paid Gemini application programming interface, or API, on October 26. Generate Content requests in those projects now begin with storage enabled. The change covers generating an answer, streaming an answer as it arrives and submitting requests in batches. The named methods are generateContent, streamGenerateContent and batchGenerateContent. Existing projects keep the settings they already have, with the announced default applying to newly created paid-tier projects.[1], [2]

The Generate Content family now follows the storage default already used by the Interactions API. Google’s change concerns the behaviour of requests in newly created paid projects. It leaves the developer with both project-wide and individual-request controls, so the setting selected for a project and the value sent with a particular call are distinct parts of the configuration.[1]

A request can override the project setting

Developers can send store=false to prevent an individual request from being stored. For a project-wide change, they can disable Generate Content API logging in the Settings panel of Google AI Studio’s Logs and Datasets dashboard. The guide says the value supplied on a request takes precedence over the project setting. A project’s default therefore does not remove control over an individual call.[1]

Google AI Studio provides the dashboard in which developers can inspect stored logs, examine model behaviour and debug applications. Sharing usage feedback is optional. The settings for Generate Content and Interactions logging can be changed independently, rather than requiring the two call families to have identical settings in every project.[1]

Logging scope includes compatible endpoints

Log storage is available for paid-tier projects. Google’s guide includes calls made through OpenAI-compatible endpoints in the supported logging scope and excludes Managed Agents calls. For Interactions, switching logging off also affects automatic storage and retrieval of conversation history unless a request explicitly overrides that choice. These controls accompany the default change without changing the settings of older projects.[1]

References

  1. News sourceGoogle AI for DevelopersGemini API stores requests by default in new paid projects↩1↩2↩3↩4↩5
  2. News sourcedailyAIGemini API stores requests by default in new paid projects↩