🎨 GIMP MCP - AI Image Generator for VS Code

Generate images from text descriptions using Stable Diffusion AI - integrated with VS Code through Model Context Protocol (MCP)

Status Python License

📖 Overview

GIMP MCP is an AI-powered image generation tool that integrates with VS Code through the Model Context Protocol. It enables you to generate images from text descriptions using Stable Diffusion, with automatic fallback to PIL-based sketches.

✨ Key Features

🤖 AI-Powered: Uses Stable Diffusion for realistic image generation
🎯 VS Code Integrated: Works seamlessly with GPT-4o Agent via MCP
⚡ Fast: Sub-second generation for sketches, ~10s for AI images
🎨 Dual Mode: AI-generated or quick PIL-based drafts
📦 Batch Processing: Generate multiple images at once
🎬 Animation Support: Create animated GIFs from text sequences
🛠️ Easy Setup: Simple bash script for quick generation

🚀 Quick Start

Prerequisites

Python 3.9+
VS Code with GPT-4o Agent
Internet connection (for AI mode)

Installation

# 1. Navigate to project directory cd /path/to/GimpMCP # 2. Create virtual environment python3 -m venv .venv # 3. Activate virtual environment source .venv/bin/activate # 4. Install dependencies pip install -r requirements.txt

Generate Your First Image

Using the bash script:

./generate_image.sh "A serene mountain landscape at sunset" "mountain_sunset"

Using Python directly:

python3 MCP/gimp-image-gen/gimp_image_gen.py \ --prompt "A serene mountain landscape at sunset" \ --output_file "output/mountain_sunset.png" \ --use-ai

📚 Documentation

QUICKSTART.md - Get started in 30 seconds
ANIMATION_GUIDE.md - Create animated GIFs from text sequences
PROMPT_EXAMPLES.md - 20+ sample prompts and best practices
INTEGRATION_SUMMARY.md - Technical implementation details
CHANGELOG.md - Version history and updates

💻 Usage

Two Generation Modes

1. AI Mode (Recommended)

Uses Stable Diffusion for realistic, detailed images:

./generate_image.sh "A futuristic cyberpunk city at night" "cyberpunk"

2. Sketch Mode

Fast PIL-based placeholders for quick drafts:

python3 MCP/gimp-image-gen/gimp_image_gen.py \ --prompt "Quick concept sketch" \ --output_file "output/sketch.png"

VS Code Integration

Ask the GPT-4o Agent directly:

"Generate an image of a vintage sports car in red, sunset lighting"

The agent will automatically invoke the tool and create your image.

Batch Generation

Generate multiple images at once:

python3 batch_generate.py output/storyboards

Animation Generation

Create animated GIFs from text sequences:

./generate_animation.sh \ "A sunrise over mountains" \ "Morning light on peaks" \ "Full daylight landscape" \ "mountain_animation"

See ANIMATION_GUIDE.md for detailed animation features.

🎨 Examples

Check PROMPT_EXAMPLES.md for 20+ categorized examples:

🚗 Vehicles & Transportation
🏠 Architecture & Buildings
👤 Characters & People
🌄 Nature & Landscapes
🎨 Abstract & Artistic
🍕 Food & Beverages
🎮 Games & Technology

Sample Outputs

Prompt	Output
"A simple 2D Volkswagen Beetle car in black and white, side view, minimalist design"	`beetle_ai.png`
"A serene Japanese garden with cherry blossoms, koi pond, stone lantern"	`japanese_garden.png`
"A futuristic cyberpunk city at night with neon lights, flying cars"	`cyberpunk_city.png`

🗂️ Project Structure

GimpMCP/ ├── .venv/ # Python virtual environment ├── .vscode/ │ └── settings.json # VS Code MCP configuration ├── MCP/ │ └── gimp-image-gen/ │ ├── manifest.json # MCP tool definition │ └── gimp_image_gen.py # Core image generator ├── output/ # Generated images ├── animations/ # Animation frames and GIFs ├── generate_image.sh # Quick generation script ├── generate_animation.sh # Animation generation script ├── generate_animation.py # Animation generator (Python) ├── batch_generate.py # Batch processing ├── requirements.txt # Python dependencies └── README.md # This file

⚙️ Configuration

AI Provider

Currently uses Pollinations.ai (free, no API key required).

To add Hugging Face support:

Get API token from Hugging Face
Edit gimp_image_gen.py:

headers["Authorization"] = "Bearer YOUR_TOKEN"

🔧 Troubleshooting

Common Issues

"Module 'PIL' not found"

source .venv/bin/activate pip install Pillow

"Permission denied: ./generate_image.sh"

chmod +x generate_image.sh

"AI generation failed"

Check internet connection
Tool automatically falls back to PIL mode
Try simplifying the prompt

Getting Help

Review QUICKSTART.md for setup issues
Check PROMPT_EXAMPLES.md for prompt tips
See INTEGRATION_SUMMARY.md for technical details

🎯 Tips for Best Results

For AI Mode

Be specific and descriptive
Include artistic style ("photorealistic", "cartoon style", etc.)
Mention lighting ("golden hour", "studio lighting")
Add quality terms ("highly detailed", "professional")

Example Good Prompts

✅ "A luxury sports car in metallic blue, side view, detailed reflections, photorealistic, studio lighting" ✅ "A cozy coffee shop interior, warm lighting, wooden furniture, people chatting, illustration style" ✅ "A majestic medieval castle on hilltop, dramatic cloudy sky, fantasy art style, epic atmosphere"

Example Poor Prompts

❌ "A car" (too vague) ❌ "Nice picture of something" (no specifics) ❌ "Image" (not descriptive)

📊 Performance

Mode	Speed	Quality	Use Case
AI Mode	~10-15s	High	Final images, presentations
Sketch Mode	<1s	Basic	Placeholders, quick drafts
Animation (5 frames)	~45s	High	Animated sequences, demos

🚀 Future Enhancements

Support for DALL·E and other AI providers
Custom image sizes and aspect ratios
Style transfer from reference images
Video format support (MP4, WebM) for animations
Advanced GIMP scripting integration
WebUI for non-technical users

📄 License

MIT License - See LICENSE for details

🙏 Acknowledgments

Stable Diffusion - AI image generation model
Pollinations.ai - Free AI inference API
PIL/Pillow - Python imaging library
Model Context Protocol - VS Code integration framework

📞 Support

For questions or issues:

Check the documentation above
Review PROMPT_EXAMPLES.md
See INTEGRATION_SUMMARY.md

Version: 3.1.0
Last Updated: November 19, 2025
Status: ✅ Production Ready

Made with ❤️ for creative workflows

GIMP MCP