Google Adds Image Generation to Gemini Based on Emails, Photos, and Calendar

Google is taking the next step toward truly personal AI: Gemini now generates images based not only on text prompts, but also on data from the Google ecosystem — from Gmail and Photos to Calendar and Drive. The new update adds image generation powered by Nano Banana 2 to the Personal Intelligence feature. The logic is straightforward: instead of writing a long prompt listing tastes, habits, and personal details, a user can write something like "create a visual of my dream home," and Gemini will automatically pull the necessary context from connected services.

If Personal Intelligence is linked to Google Photos, the model can use not only general preferences but also real photos and tags like "family" or "pets" to generate more precise and personal images.

For Google, this represents an important expansion of the strategy launched in January 2026 alongside Personal Intelligence. Until now, the feature has mostly worked in text scenarios: helping find travel details in email, gathering calendar plans, accounting for purchase history, and making Gemini responses less templated. Now the same idea is being extended to visual format. Personalization is becoming not an add-on to the chat, but part of the content generation mechanism itself. This is particularly important given competition, where users already expect AI to deliver not just beautiful pictures, but results that account for their habits, interests, and surroundings from the first request.

Technologically, Google is betting on Nano Banana — an embedded family of image generators for Gemini, rather than the Imagen line, which is more oriented toward standalone visual and professional tasks. This line already has several versions: basic Nano Banana for quick everyday scenarios, Nano Banana 2 with faster iterations and expanded capabilities, and Nano Banana Pro, which leverages the full reasoning power of Gemini 3 Pro. The key wager here is that the image is created not by an external module, but within the same system that understands text, conversation context, and user background tied to the account.

Google specifically emphasizes transparency and control. If Gemini picks the wrong reference or gets details wrong, the result can be refined with plain text or by manually adding a different photo via the "+" button. For personalized images, a Sources section has also appeared: it shows which data or photos were used as the foundation. The company claims it does not train models directly on Google Photos' private library, and Personal Intelligence connection remains voluntary and user-configurable. This is a crucial point: the deeper AI relies on personal data, the higher the sensitivity to errors, privacy, and trust.

The rollout is happening in phases. On April 16, 2026, Google announced that the feature is beginning to appear for Google AI Plus, Pro, and Ultra subscribers in the United States. Later, it promises to add it to Gemini in Chrome on desktop and expand to other audiences. However, the launch is not happening everywhere: Europe was not included in the first wave, and Google's reference materials list the European Economic Area, United Kingdom, and Switzerland among unavailable regions. For the company, this is a typical cautious launch scenario, especially when products are closely tied to personal data.

The main takeaway is simple: the next battle in the AI market is no longer only about generation quality, but about the depth of personal context. Google has an almost unique position because it already has email, photos, calendar, search, maps, and video history all within a single account. If the company can maintain the balance between convenience and privacy, Gemini will gain a very strong advantage: less manual description, more useful results from the first request.