Google just made it significantly less annoying to put yourself in AI-generated content. Gemini now supports persistent personal avatars: record your face and voice once, and you can drop yourself into any image or video scene without uploading a new selfie every time. The feature is rolling out today to Gemini app users in select countries, and it ties directly into Nano Banana, Google's viral image generation model.
What is Nano Banana, exactly?
The model powering Nano Banana is Gemini 2.5 Flash Image, Google's current-generation multimodal image model. The "Nano Banana" nickname was popularized by creators and AI content communities online as shorthand for Google's Gemini Flash Image model. The name stuck because the model punches well above its weight class for a free-to-access tool. It has since become the backbone of Gemini's entire personalized image stack.
Nano Banana refers to four distinct models available in the Gemini API, including Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image), the fastest and cheapest option engineered for velocity and scale. Nano Banana 2 pulls from the Gemini model's real-world knowledge base and is powered by real-time information and images from web search to more accurately render specific subjects.
The selfie problem it solves
Every time you wanted to generate an image of yourself with a previous AI tool, you had to manually upload a photo, describe your appearance in the prompt, and hope the model kept your likeness consistent. That friction adds up fast, especially if you're iterating across multiple styles or scenes.
Don't miss what's next in AI
Join 300,000+ engineers and researchers who get the signal, not the noise.
- Full access to in-depth AI research breakdowns
- Be the first to know what's trending before it hits mainstream
- Daily curated papers, repos, and industry moves