Nano Banana Pro Release Notes

Google DeepMind has released Nano Banana Pro (also known as Gemini 3 Pro Image), a state-of-the-art image generation and editing model. Built on the Gemini 3 Pro architecture, the model integrates advanced reasoning and real-world knowledge to provide higher precision in visual synthesis and editing than its predecessor, Nano Banana.

Enhanced Reasoning and World Knowledge

Nano Banana Pro utilizes Gemini 3's reasoning capabilities to generate context-rich visuals that function as educational tools and data representations. The model can create accurate infographics and diagrams based on user-provided content or factual real-world data.

Integration with Google Search allows the model to visualize real-time information, such as current weather or sports updates, and generate quick snapshots for recipes. This shift moves the model beyond simple aesthetic generation toward the creation of helpful, information-dense content.

Advanced Text Rendering and Multilingual Support

Nano Banana Pro improves the rendering of legible, accurate text within images, supporting both short taglines and long paragraphs. The model's understanding of depth and nuance allows for a wider variety of fonts, textures, and calligraphy in mockups and posters.

Due to Gemini 3's multilingual reasoning, the model can generate text in multiple languages and localize content for international scaling.

Professional Creative Controls and Fidelity

Nano Banana Pro introduces several high-fidelity capabilities designed for professional workflows:

  • Consistency and Blending: The model can blend up to 14 images and maintain the resemblance of up to 5 people, enabling the transformation of sketches into products or blueprints into photorealistic 3D structures.
  • Studio-Quality Editing: Users have access to localized editing for selecting and refining specific image parts. Creative controls include adjusting camera angles, changing focus, applying color grading, and transforming scene lighting (e.g., creating bokeh effects or shifting day to night).
  • Resolution and Aspect Ratios: The model supports various aspect ratios and offers outputs in 2K and 4K resolution.

Availability and Integration

Nano Banana Pro is being deployed across the Google ecosystem with different tiers of access:

  • Consumers and Students: Available globally in the Gemini app via the 'Thinking' model. Free-tier users have limited quotas before reverting to the original Nano Banana model; Google AI Plus, Pro, and Ultra subscribers have higher quotas. In the U.S., it is available in AI Mode in Search for Pro and Ultra subscribers, and globally in NotebookLM for subscribers.
  • Professionals: Integrated into Google Ads globally, and rolling out to Google Slides and Vids for Workspace customers.
  • Developers and Enterprise: Rolling out in the Gemini API, Google AI Studio, and Google Antigravity for UX layouts. Vertex AI is available for scaled enterprise creation, with Gemini Enterprise coming soon.
  • Creatives: Rolling out to Google AI Ultra subscribers in Flow, Google's AI filmmaking tool.

AI Identification and Transparency

To ensure transparency, Google employs a multi-layered approach to identify AI-generated content:

  • SynthID: All media generated by Google's tools include an imperceptible digital watermark. Users can now upload images to the Gemini app to verify if they were generated by Google AI using English language prompts.
  • Visible Watermarks: A "Gemini sparkle" watermark is maintained for images generated by free and Google AI Pro tier users.
  • Professional Exceptions: Visible watermarks are removed for Google AI Ultra subscribers and users of the Google AI Studio developer tool to provide a clean canvas for professional work.

Sources