generate-image

To generate images from text, you provide a detailed prompt describing your desired visual content. The Skill uses advanced AI models to perform text-to-image generation, creating photorealistic or stylized art based on your specific instructions.

Yes, you can edit existing images by describing desired changes in natural language. The image-to-image feature processes your commands to modify visual content, allowing you to showcase different variations or update product photos without manual editing.

Yes, you need a GEMINI_API_KEY and the uv package manager to execute the text-to-image generation scripts. These dependencies are required to authenticate requests and run the underlying AI models for processing your visual content.

Multi-reference composition combines elements from multiple existing images into a new composition. This technique allows you to merge distinct visual components, such as combining a specific product with a different background, into a cohesive generated image.

Yes, the Skill supports configurable aspect ratios and resolutions for AI art generation. You can specify these parameters alongside choosing between different models, such as Nano Banana or Nano Banana Pro, to achieve precise text rendering and photorealism.

Natural language image editing relies heavily on prompt clarity and model interpretation. Complex or highly abstract instructions might not translate perfectly into the desired visual changes, requiring iterative prompting to achieve precise text rendering or specific photorealistic details.