gemini-nano-banana-tool

gemini-nano-banana-tool 🍌

Logo generated with Gemini Pro model (~$0.22 USD)

Gemini Nano Banana Tool - Professional CLI for generating, editing, and manipulating images using Google's Gemini and Imagen 4 image generation models

Table of Contents

• About
  • What is Nano Banana?
  • Why This CLI?
• Use Cases
• Features
• Installation
• Configuration
• Usage
  • Promptgen Command
  • Generate Command
  • Generate Conversation Command
  • List Commands
• Library Usage
• Resources
• Development
• License

About

What is Nano Banana?

This tool supports Google's image generation models across two families:

Gemini Models (Nano Banana)

Nano Banana (gemini-2.5-flash-image) - Fast image generation with fixed ~1024p resolution (up to 3 reference images) Nano Banana 2 (gemini-3-pro-image-preview) - Advanced pro model with 1K/2K/4K resolution support (up to 14 reference images)

Features:

• 🎨 High-quality text-to-image generation
• 🖼️ Image editing with natural language prompts
• 🔄 Multi-image composition (3 images for Flash, 14 images for Pro)
• 📐 Multiple aspect ratios (1:1, 16:9, 9:16, and more)
• 🎯 Variable resolution (Pro model: 1K/2K/4K quality levels)
• 🎭 Style transfer and artistic rendering

Imagen 4 Models

Imagen 4 Fast (imagen-4.0-fast-generate-001) - Fastest generation, cost-effective ($0.02/image) Imagen 4 (imagen-4.0-generate-001) - Balanced quality and speed ($0.04/image) Imagen 4 Ultra (imagen-4.0-ultra-generate-001) - Highest quality photorealism ($0.06/image)

Features:

• 📸 Photorealistic image generation
• ⚡ Fast generation times (especially Fast variant)
• 💰 Simple per-image pricing
• 🎯 High-quality outputs across all variants

All models provide:

• ✨ Built-in SynthID watermarking for authenticity
• 📐 Multiple aspect ratios (1:1, 16:9, 9:16, 4:3, 3:4, and more)

Learn more:

• Gemini Image Generation Documentation
• Imagen 4 on Vertex AI

Why This CLI?

This tool provides a professional, agent-friendly CLI for Gemini image generation with:

• 🤖 Agent-Friendly Design: Structured commands and error messages enable AI agents (like Claude Code) to reason and act effectively in ReAct loops
• 🔧 Composable Architecture: JSON output to stdout, logs to stderr—perfect for pipes and automation
• 📦 Reusable Building Blocks: Commands serve as building blocks for skills, MCP servers, shell scripts, and workflows
• 🛡️ Type-Safe & Reliable: Comprehensive type hints and mypy strict mode ensure predictable behavior in automated systems
• 📚 Rich Documentation: Extensive help messages and error handling guide both humans and agents
• 🎯 Dual-Mode Operation: Use as CLI tool or import as Python library

Use Cases

• 🎨 Creative Content Generation - Generate marketing visuals, social media content, concept art
• 🖼️ Image Editing & Enhancement - Remove objects, change backgrounds, apply style transfers
• 🔄 Multi-Image Composition - Combine multiple images for fashion e-commerce, product visualization
• 📐 Multi-Format Output - Generate images for various platforms (Instagram, YouTube, TikTok, etc.)
• 🤖 AI Agent Integration - Build autonomous image generation workflows with Claude Code
• 🔁 Batch Processing - Script generation pipelines with shell loops and automation
• 🧪 Rapid Prototyping - Quick visual mockups and design iterations

Features