Google DeepMind presenta Nano Banana Pro con renderizado de texto avanzado y edición visual de calidad de estudio

In Brief

Google DeepMind has launched Nano Banana Pro image generation and editing model, offering advanced reasoning, enhanced text rendering, multi-element composition, and more.

Google DeepMind, AI division of technology company Google, announced the release of Nano Banana Pro, or Gemini 3 Pro Image, a next-generation model for image generation and editing. Developed on top of Gemini 3 Pro, Nano Banana Pro leverages the model’s advanced reasoning abilities and real-world knowledge to produce more accurate visual representations.

The model is designed to support a wide range of visual tasks, including conceptual design, prototype creation, infographic development, and the conversion of handwritten content into structured diagrams. Nano Banana Pro generates context-aware visuals informed by enhanced reasoning, external information, and real-time data. By incorporating Gemini 3’s analytical capabilities, the model can produce informative educational graphics, detailed explainers, and accurate illustrations based on user-provided material or verified information from the real world. It can also draw on Google Search to visualize items such as recipes, live weather updates, or sports information.

Nano Banana Pro further improves the rendering of text within images, providing greater accuracy, clarity, and legibility across multiple languages. The system supports the creation of visuals containing text ranging from short labels to longer passages, while better understanding stylistic nuance, depth, and layout. This enables more precise text integration in mockups, posters, and other creative material, and allows content to be localized or translated for broader use.

The model also enhances visual fidelity and creative flexibility. Its design makes it possible to combine multiple elements, incorporate up to fourteen images, and maintain consistent representations of as many as five individuals within a composition. Tasks such as converting sketches into product concepts or transforming blueprints into detailed 3D-style renderings benefit from its improved coherence. Nano Banana Pro introduces expanded creative controls, allowing users to adjust camera perspectives, modify focus, refine specific regions of an image, apply complex color grading, and alter lighting conditions. Outputs can be generated in a variety of aspect ratios, with support for 2K and 4K resolution to meet both digital and print-quality requirements

Nano Banana Pro Now Available Across Consumer, Enterprise, And Creative Tools

Users across Google’s products and services can now choose between the original Nano Banana model, designed for quick and simple edits, and Nano Banana Pro, intended for more advanced compositions that require high-quality, visually refined output. The enhanced model is gradually becoming available to consumers and students through the Gemini app by selecting “Create images” with the “Thinking” model. Free-tier users receive limited access before transitioning back to the original version, while subscribers to Google AI Plus, Pro, and Ultra plans receive expanded quotas. In the United States, Nano Banana Pro is accessible in AI Mode for Search for Google AI Pro and Ultra subscribers, and it is also offered to NotebookLM subscribers worldwide.

For professional users, Google Ads is integrating Nano Banana Pro to support more advanced creative and editing capabilities for advertisers globally. The model is also rolling out to Google Workspace applications, including Google Slides and Vids. Developers and enterprise clients will see Nano Banana Pro introduced through the Gemini API, Google AI Studio, and Google Antigravity to support richer layout and mockup generation. Enterprises can begin using the model within Vertex AI for large-scale creation, with additional availability planned for Gemini Enterprise. Creative professionals will gain access through Google AI Ultra subscriptions in Flow, the company’s AI filmmaking tool, enabling greater precision and control when shaping scenes and visual frames.

Google continues to emphasize transparency around AI-generated media. All images produced through its tools include an imperceptible SynthID digital watermark. Users can now upload an image into the Gemini application and directly inquire whether it was generated by Google AI via SynthID verification, with support for audio and video planned in the future. Alongside SynthID, a visible Gemini watermark will appear on images created by free-tier and Google AI Pro users to support clearer identification. For professional workflows requiring unobstructed visuals, the visible watermark will be removed from images produced by Google AI Ultra subscribers and those generated within the Google AI Studio development environment