nano-image-generator

To generate app icons and marketing banners from text prompts, use a script-driven workflow that automates image creation from natural language, reducing design cycle times for UI graphics and visual assets.

Yes, you can use reference images for style transfer and character consistency. The image generation process accepts up to 14 reference images to apply specific visual styles and maintain character consistency across generated outputs.

Prompt-based image generation supports multiple aspect ratios including 1:1, 16:9, and 9:16, alongside multiple resolutions ranging from 1K to 2K and 4K for flexible output of visual assets.

Yes, you need a Gemini API key to generate UI graphics and social media images. The image generation workflow uses Nano Banana Pro (Gemini 3 Pro Preview) to create visual assets from your natural language prompts.

You can integrate image generation into automated pipelines by running the script-driven workflow ad-hoc or within pipelines. Execute the generate_image.py script with a prompt and an output path to generate visual assets.

The best way to create logos and illustrations without manual design work is using prompt-based image generation. This approach automates visual asset creation from natural language descriptions, significantly reducing design cycle times.