Why does the skill use uppercase 'K' for resolutions?

Google's image generation API is case-sensitive; using lowercase '4k' results in a request failure. This skill automatically handles the correct '4K', '2K', and '1K' formatting.

Which Gemini models are supported for image generation?

The skill supports Gemini 3 Pro Image Preview for high-resolution 4K images and Gemini 2.5 Flash Image for fast iterations and simple edits.

Can I generate images with legible text?

Yes, Gemini 3 Pro Image offers approximately 94% legibility for text in images at 4K resolution, making it significantly more reliable than DALL-E 3 for infographics.

Does this skill handle the latest Google AI SDK?

Yes, it enforces the mandatory migration from the deprecated @google/generative-ai package to the new @google/genai SDK required for all projects after November 2025.

How many reference images can I use for character consistency?

You can use up to 14 reference images total, but you must limit human subjects to 5 or fewer to maintain predictable character consistency.

Gemini Native Image Generation

Name: Gemini Native Image Generation
Author: evolv3ai

byevolv3ai

0•

设计工具

Generates and edits high-resolution website images, infographics, and hero banners using the latest Gemini Native Image models.

This skill provides a standardized framework for Claude to generate, edit, and manipulate website images using Gemini 3 Pro and Gemini 2.5 Flash models. It specializes in creating production-ready assets like hero banners and service cards with legible text, supports multi-turn editing for style consistency, and includes specific logic for Australian-themed imagery. By implementing critical SDK migration patterns and addressing known API limitations like resolution case-sensitivity and human reference limits, it ensures reliable image generation for frontend development workflows.

主要功能

01Support for 4K resolution and ultra-wide 21:9 aspect ratios

02Reference image support with character consistency logic

03High-legibility text rendering in images and infographics

040 GitHub stars

05Automated SDK migration patterns for Google GenAI integration

06Multi-turn editing for style transfer and element manipulation

使用场景

01Generating branded infographics and diagrams with readable text overlays

02Creating custom website hero banners and service cards when stock photos are unavailable

03Performing AI-driven image edits, such as changing colors or adding/removing scene elements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add evolv3ai/claude-skills-archive image-gen

For use in Claude.ai and ChatGPT

Download Skill