Easy as pie: Google's Gemini uses your memories for AI photos that feel personal

Google is adding personalized image generation to Gemini and Nano Banana 2.

The feature can use connected Google Photos, photo labels, and signals from activity such as YouTube and Chrome to create more tailored images.

Users can get results that reflect their preferences and memories from just a short prompt.

The rollout is optional, and photos are not used to train the model.

It is launching first for U.S. AI Pro, Plus, and Ultra subscribers.