Nano Banana Pro (Gemini 3 Pro Image) Release

Google DeepMind has released Nano Banana Pro (Gemini 3 Pro Image), a high-fidelity image generation and editing model that provides developers with studio-quality image generation and advanced creative control. Built on Gemini 3 Pro, this model is designed for professional production, offering higher accuracy in text rendering, robust world knowledge, and integration with Google Search for grounded, data-driven outputs.

Professional-Grade Fidelity and Creative Control

Gemini 3 Pro Image enables precise control over image composition and physics, including lighting, camera settings, focus, and color grading. To meet professional production standards, the model supports 2K and 4K resolutions.

Key capabilities for cohesive asset creation include:

  • Reference Integration: The model can combine diverse elements such as logos, product images, and references into single advertisements.
  • Consistency: It can maintain consistent resemblance for up to five individuals.
  • Input Blending: The model supports integrating up to six high-fidelity shots or blending up to fourteen standard inputs into a single polished output.

Advanced Text Rendering and Localization

Gemini 3 Pro Image provides a significant improvement in text rendering over the previous 2.5 Flash Image model, producing clear and accurate text integrated directly into images. This capability makes it suitable for creating functional assets such as marketing collateral and educational content.

Beyond generation, the model supports localization logic through image-to-image generation. By grasping the semantic context of an image, it can change the language of elements like signs, menus, or documents while preserving the original artistic style and layout.

Grounding and World Knowledge

To produce more factual assets, Gemini 3 Pro Image leverages a vast knowledge base and can be connected to real-time web content via grounding with Google Search. This integration is specifically useful for applications requiring high precision, such as historical maps or biological diagrams.

Ecosystem Integration and Safety

Gemini 3 Pro Image is being integrated into the developer ecosystem through several channels:

  • Google Antigravity: Coding agents on this agentic development platform can use the model to generate UI mockups and visual assets before writing code.
  • Creative Platforms: The model is being integrated into Adobe and Figma.
  • Developer Access: The model is available in paid preview via the Gemini API in Google AI Studio and Vertex AI for enterprises.

To ensure the provenance of AI-generated media, every image created or edited with Gemini 3 Pro Image includes integrated SynthID digital watermarks.

Model Comparison: Nano Banana vs. Nano Banana Pro

Developers can choose between two models based on their specific needs:

  • Gemini 2.5 Flash Image (Nano Banana): Optimized for faster, lower-cost image generation.
  • Gemini 3 Pro Image (Nano Banana Pro): Optimized for higher quality image generation, with higher associated costs and latency.

Sources