Nano Banana 2 Skill is a command-line interface (CLI) tool for AI-powered image generation, leveraging Google's Gemini models. It targets developers and power users seeking advanced image manipulation capabilities, offering features like transparent asset generation, style transfer, and integration with AI assistants like Claude, streamlining complex visual content creation workflows.

How It Works

The CLI interfaces with Gemini models, supporting Gemini 3.1 Flash (default, fast, cost-effective) and Gemini 3 Pro (higher quality), along with any specified Gemini model ID. It enables multi-resolution outputs from 512px to 4K and customizable aspect ratios. A key innovation is its broadcast-grade green screen transparency feature, which uses a prompt-based green screen generation, followed by FFmpeg for colorkeying and despilling, and ImageMagick for final trimming, ensuring clean, transparent assets.

Quick Start & Requirements

• Installation: Requires Bun, FFmpeg, and ImageMagick. Clone the repository, install dependencies with bun install, and link globally using bun link.
• API Key: Configure your Gemini API key by creating a ~/.nano-banana/.env file with GEMINI_API_KEY=your_key_here. API keys can be obtained from Google AI Studio.
• Claude Code Skill: The tool also functions as a Claude Code skill, automating setup and command generation within Claude's environment.
• Documentation: Usage examples and detailed options are provided in the README.

Highlighted Details

• Supports Gemini 3.1 Flash (speed/cost) and Gemini 3 Pro (quality) models.
• Generates images at resolutions from 512px up to 4K with configurable aspect ratios.
• Enables advanced image manipulation via reference images for editing, style transfer, and precise dimension control.
• Features an automated green screen to transparent background pipeline using FFmpeg and ImageMagick.
• Includes built-in cost tracking for generations, accessible via nano-banana --costs.
• Integrates seamlessly as a Claude Code skill for AI-assisted command generation.

Maintenance & Community

No specific details regarding maintainers, community channels (like Discord/Slack), or roadmap were present in the provided documentation.

Licensing & Compatibility

The project is released under the MIT license, which generally permits commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

Successful operation is contingent on the correct installation and configuration of external dependencies: Bun, FFmpeg, and ImageMagick. Users must also obtain and manage their own Gemini API keys. The advanced transparency features require these specific system-level tools to be installed and accessible.