Google Makes Gemini Personalized Image Generation Free for US Users

Google has made Gemini's personalized AI image generation feature free for eligible users in the United States, dropping the paywall that previously restricted it to Plus, Pro, and Ultra subscribers since its launch in April 2026. The expansion, announced on June 29, 2026, applies to any US user aged 13 or older.

Personalized image generation, known internally as Nano Banana, uses data from connected Google apps such as Gmail, Calendar, and Photos to tailor images based on an individual's interests and habits. A user asking for a birthday party invitation image might get a design incorporating their child's favorite cartoon from a recent email, for instance.

Feature Rollout and Eligibility

The feature is now enabled by default for free users in the US who have opted into Google's personalization settings. Key details:

• Users must have a Google account with personalization turned on in Gemini settings.
• Images are generated using a fine-tuned version of Google's Imagen model, which applies the user's personalized style and subject preferences.
• Google says images are not used to train the model and can be deleted at any time from the user's Gemini activity history.
• The feature supports multiple outputs, allowing users to pick from variations before saving or sharing.
• Eligibility is limited to US accounts; no timeline for international expansion has been provided.

Competitive Pressure and Data Privacy

Google's decision to unbundle Nano Banana from paid tiers follows complaints from users who criticized the paywalled personalization as limiting the appeal of its free Gemini tier. Competitors like OpenAI's ChatGPT Plus and Microsoft's Copilot Pro offer similar image generation capabilities only within their paid subscriptions. By contrast, Google is betting that a free personalized generation tool will drive more users to engage with Gemini, increasing ad exposure and potential upgrades to other Google One benefits.

The move also raises privacy questions. Google has emphasized that the personalization happens on-device where possible, with server-side processing only for the image generation step. The company states that user data is encrypted and that the generated images do not retain personal identifiers. Privacy advocates, however, have flagged that the opt-in default could cause some users to unknowingly share sensitive context for image requests.

Google plans to expand personalized image generation to other regions later in 2026. The company is also testing personalized text responses based on user profile data, a feature it describes as a step toward more context-aware AI assistants. For now, US users can create unlimited personalized images at no cost, a shift that may pressure rivals to reconsider their own pricing strategies for generative AI tools.