Google has launched Nano Banana Pro, its new image generation and editing model built on the latest Gemini 3 Pro. Nano Banana Pro builds on the original Nano Banana, which Google released in August as a Gemini 2.5 Flash Image model for basic photo restoration and figurine generation.
The new version adds reasoning capabilities and real-world knowledge from Google Search, enabling users to create context-rich visuals such as infographics, guides, and data visualizations grounded in factual information.
One of the best improvements is in text rendering. The model generates images with correct, legible text in multiple languages.
For complex projects, Nano Banana Pro combines up to 14 elements while maintaining the consistency of up to 5 people, ideal for editorial shoots, storyboards, and branded content. The model also supports high-fidelity editing with studio-quality controls that let users adjust camera angles, apply color grading, control lighting (day-to-night and bokeh effects), and modify aspect ratios for different platforms. Output reaches up to 4K resolution.
Google is rolling out Nano Banana Pro across multiple services. Free and paid Gemini users get access through the app; professionals gain it through Google Ads, Slides, and Vids; developers access it via Gemini API and AI Studio; and filmmakers receive it in Flow for video editing.
Free and Pro tier users see a visible Gemini watermark, while Ultra subscribers and developers work without visible markers. Users can now verify Google-generated images directly in the Gemini app.