claude-code-image

claude-code-image

Generate and edit images using the latest models all within Claude Code.

Install

/plugin marketplace add jshchnz/claude-code-image
/plugin install image@claude-code-image

Requires Claude Code v1.0.33+

Then configure your API key:

"Set up image generation" or /image:configure

Quick Start

After installing, simply ask Claude to create images:

"Create an image of a serene Japanese garden at sunset"

"Generate a logo for a tech startup called 'NovaSpark'"

"Make me a cute cartoon fox sticker"

Claude will handle the rest - selecting the right service, generating the image, and saving it for you.

Examples

Creating Images

Landscapes & Scenes:

"Create an image of a cyberpunk city at night with neon lights and flying cars"

Logos & Icons:

"Generate a minimalist logo for a coffee shop called 'Bean There'"

Characters & Art:

"Draw a friendly robot mascot in a cartoon style"

Stickers (auto-transparent):

"Make a kawaii cat sticker with sparkles"

Editing Images

Adding Elements:

"Here's my photo [attach image]. Add sunglasses to me"

Style Transfer:

"Take this landscape photo and make it look like a Van Gogh painting"

Modifications:

"Edit this image and change the background to a beach"

Slash Commands (Advanced)

For explicit control, you can also use slash commands:

Command	Description
/image:generate <prompt>	Generate with specific prompt
/image:edit	Interactive image editing
/image:configure	Set up API keys and defaults
/image:models	View available AI models
/image:history	See recent generations
/image:status	Check configuration

Available Models

GPT Image (OpenAI)

Model	Speed	Quality	Cost
gpt-image-1.5	Medium	Best	~$0.02-0.08/image
gpt-image-1	Medium	High	~$0.02-0.08/image
gpt-image-1-mini	Fast	Good	~$0.005-0.02/image

Nanobanana (Google Gemini)

Model	Speed	Quality	Notes
gemini-2.5-flash-image	Fast	Good	Free tier available
gemini-3-pro-image-preview	Slower	Best	Up to 4K resolution

Configuration

API Keys

Set up your API keys as environment variables:

# OpenAI (for GPT Image)
export OPENAI_API_KEY='your-openai-api-key'

# Google (for Nanobanana)
export GEMINI_API_KEY='your-gemini-api-key'

Get your API keys:

• OpenAI: https://platform.openai.com/api-keys
• Gemini: https://aistudio.google.com/apikey

User Preferences

Run /image:configure or say "configure image generation" to set up:

• Default service (GPT Image or Nanobanana)
• Default model for each service
• Output directory
• Default size, quality, and format

Configuration is saved to ~/.claude-code-image/config.json.

Output Options

Size

• square - 1024x1024
• portrait - 1024x1536
• landscape - 1536x1024

Quality

• auto - Let the AI decide
• low - Fastest
• medium - Balanced
• high - Best detail

Format

• png - Best for transparency
• jpeg - Smaller file size
• webp - Modern format

Auto-Transparency

The plugin automatically enables transparent backgrounds when your prompt contains keywords like:

• sticker, icon, sprite, logo
• cutout, transparent
• asset, emoji

File Locations

Path	Purpose
~/.claude-code-image/config.json	User preferences
~/.claude-code-image/history.json	Generation history
~/Pictures/claude-code-image/	Default output directory

Requirements

• Claude Code CLI
• curl (for API requests)
• base64 (for encoding/decoding)
• python3 (for JSON parsing in Gemini scripts)
• API key for at least one service (OpenAI or Google)

License

MIT License - see LICENSE for details.

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

Support

• Report issues: https://github.com/jshchnz/claude-code-image/issues
• Discussions: https://github.com/jshchnz/claude-code-image/discussions

Acknowledgments

• OpenAI for the GPT Image API
• Google for the Gemini API
• Anthropic for Claude Code