Gemini no longer needs your prompts: AI will browse your photos and make you a star

  • The Gemini app has learned to generate images using photos from Google Photos
  • The feature builds on the Nano Banana 2 model and works with tagged people and pets in the library
  • The new feature is now gradually rolling out to Google AI Plus, Pro, and Ultra subscribers in the USA

Sdílejte:
Adam Kurfürst
Adam Kurfürst
17. 4. 2026 14:30
Advertisement

Image generation is now one of the most popular ways people let AI assistants intervene in their creative lives. The catch is that if you really want yourself or someone close to you in the result, you usually can’t do without a detailed prompt and manually uploading a reference photo. Google is now trying to radically simplify this in the Gemini app – AI will pull the necessary assets directly from your Google Photos.

Your family in the starring role, and you don’t have to lift a finger

The new feature falls under the overarching concept of Personal Intelligence, which Google has been developing in Gemini for a long time (unfortunately, it still hasn’t reached Europe). It connects linked Google services with the generative model Nano Banana 2, so that the app understands the context of your life. For images, this practically means that you just need to write a simple prompt like “create an image of me and my family doing our favorite activity.” Gemini will find your faces in Google Photos based on existing people and animal tags and project them into the resulting image.

Similarly, you can ask for a portrait in the style of watercolor, charcoal drawing, or oil painting, and Gemini will generate it with you in the starring role, without the need to upload anything. If the AI uses the wrong photo or the result misses the mark, you can manually select a different reference via the “+” button, specify what went wrong, and let Gemini try again. The Sources button also reveals which specific image from your library the model just used.

Privacy solved, availability not yet

With such a sensitive feature, privacy is naturally the first question. Google explicitly states in the announcement that it does not train its models directly on your private Google Photos library. Training only occurs on limited data, such as the prompts themselves and the model’s responses. The connection of Google services with Gemini is also optional and can be turned off at any time in the settings – so for users who don’t want AI to browse their gallery, it’s still a choice, not a necessity.

Availability is a different story. The new feature is gradually rolling out over the next few days only to paying subscribers of Google AI Plus, Pro, and Ultra plans in the USA. Google promises its expansion to Gemini in Chrome on desktops and a later release for other users, but has not published a specific timeline for Europe or specifically the Czech Republic. However, it probably won’t arrive here until the company expands Personal Intelligence to us, which may take a while.

Would you let Gemini access your Google Photos directly for inspiration?

Source: Google

About the author

Adam Kurfürst

Adam studuje na gymnáziu a technologické žurnalistice se věnuje od svých 14 let. Pakliže pomineme jeho vášeň pro chytré telefony, tablety a příslušenství, rád se… More about the author

Adam Kurfürst
Sdílejte: