Latest Google Vertex AI optimized for faster image generation
# Nano Banana 2 (G) Versatile AI model for high-quality creative and analytical tasks. ## Overview Nano Banana 2 (G) represents the next evolution in high-efficiency image generation, powered by the advanced Gemini 3.1 Flash Image architecture. Designed specifically for marketing professionals and creative agencies who require a balance between rapid iteration and high-fidelity output, this model excels at transforming abstract concepts into polished visual assets. By leveraging a streamlined inference pipeline, Nano Banana 2 (G) provides a robust platform for both generative art and precise image editing, ensuring that your creative vision is realized with minimal latency. What sets this model apart is its deep integration with Google’s latest multimodal reasoning capabilities. Unlike standard generative models, Nano Banana 2 (G) utilizes a sophisticated 'thinking' layer that allows it to interpret complex prompts with greater contextual awareness. Whether you are generating high-resolution campaign imagery or performing nuanced edits on existing visual assets, the model maintains consistent style adherence and structural integrity, making it an indispensable tool for fast-paced advertising workflows. ## Use Cases * **Rapid Social Media Content:** Generate high-quality, platform-ready imagery for Instagram, LinkedIn, and TikTok campaigns in seconds. * **Dynamic Ad Variations:** Quickly iterate on ad creatives by adjusting aspect ratios and styles to suit different display network requirements. * **Brand Asset Localization:** Use the image editing capabilities to swap product placements or adjust background environments while maintaining brand consistency. * **Concept Visualization:** Transform rough sketches or mood board descriptions into photorealistic mockups for client presentations. * **E-commerce Product Photography:** Enhance product shots by generating professional-grade backgrounds or lifestyle settings without the need for physical studio time. * **Marketing Material Prototyping:** Create consistent visual themes for email newsletters, landing pages, and digital brochures. ## Parameters Nano Banana 2 (G) offers a comprehensive suite of configuration options to fine-tune your creative output. Please refer to the Input Parameters section in the Pixloop interface to adjust settings such as temperature, aspect ratio, and grounding, ensuring your generation aligns perfectly with your project requirements. ## Tips for Best Results * **Leverage Grounding:** Enable the grounding feature when your prompt involves current events, specific trending aesthetics, or real-world locations to ensure the model pulls from the most relevant, up-to-date visual context. * **Iterate with Temperature:** Use a lower temperature (0.2–0.5) for precise, brand-compliant assets, and increase to 1.2+ when you need the model to explore more creative, abstract, or experimental visual directions. * **Utilize Thinking Levels:** Set the thinking_level to 'HIGH' for complex, multi-layered prompts that require the model to synthesize multiple visual elements or specific stylistic constraints. * **Optimize Aspect Ratios:** Always select the aspect ratio that matches your final delivery platform (e.g., 9:16 for Stories, 16:9 for YouTube) to avoid cropping artifacts and maintain composition quality. * **Reference Images:** When performing image edits, provide high-quality input images to guide the model’s understanding of lighting, texture, and perspective. ## About Nano Banana 2 (G) is powered by the Vertex AI platform and utilizes the Gemini 3.1 Flash Image model lineage. Developed as part of Google’s ongoing commitment to high-performance, low-latency multimodal AI, this model is optimized for creative professionals who demand both speed and precision. It builds upon the foundational research of the Gemini series, offering a specialized interface for Pixloop users to harness state-of-the-art generative technology in a production-ready environment.
Model input reference derived from preset schema.
| Name | Type | Required | Default | Description |
|---|---|---|---|---|
| top_p | number | No | 0.95 | Cumulative probability threshold for token selection (0.0-1.0) |
| prompt | string | Yes | null | Text description of the image you want to generate or edit |
| grounding | boolean | No | false | Enable Google Search grounding for fresher web context. |
| image_size | select | No | "1K" | Output image resolution (Vertex imageConfig.imageSize). |
| temperature | number | No | 1 | Controls randomness in generation (0.0 = deterministic, 2.0 = maximum creativity) |
| aspect_ratio | select | No | "1:1" | Aspect ratio of generated images |
| input_images | image | No | [] | Optional reference images for editing or style transfer (up to 10 images) |
| output_format | select | No | "png" | Requested output format for generated images |
| thinking_level | select | No | "MINIMAL" | Reasoning intensity for Gemini 3.1 Flash Image. |
| candidate_count | select | No | "1" | Number of image variations to generate (fixed at 1 per request) |