Gemini gains Nano Banana Pro image generation
By Steven Van ·
The Gemini 3 Pro-based model blends up to 14 images, keeps 5 people consistent, and outputs up to 4K with SynthID watermarks.
Nano Banana Pro, Google DeepMind's new image generation and editing model, is now built into Gemini, along with Google Ads, Workspace and Google AI Studio. It runs on Gemini 3 Pro, and Google says it renders legible text more accurately than before, including in multiple languages, which makes it usable for mockups, posters and localised content.
The model can blend up to 14 input images into one composition while keeping up to 5 people consistent, so sketches can be turned into products or blueprints into photorealistic structures. Output goes up to 4K resolution, and images carry SynthID watermarks. Nano Banana Pro can also pull in real-time information via Search grounding, for things like current weather or sports data, to build infographics and diagrams.
Full details are in Google's announcement.
