OpenAI Image Generator

MCP server that wraps OpenAI's gpt-image-2 image generation API.

Tools

Tool	Description
generate_image	Generate images from text prompts using gpt-image-2
list_generated_images	List all previously generated images in the output folder
load_image	Load and display a saved image inline

Setup

1. Build

npm install
npm run build

2. Configuration

Environment variables:

Variable	Default	Description
OPENAI_API_KEY	(required)	Your OpenAI API key
OPENAI_BASE_URL	(optional)	Custom base URL (e.g. a self-hosted proxy); defaults to official OpenAI endpoint
IMAGE_OUTPUT_DIR	./generated_images	Where to save generated images

3. Register with Claude Desktop / Kiro CLI

Add to your MCP config (e.g. ~/.claude/mcp.json or Claude Desktop claude_desktop_config.json):

{
  "mcpServers": {
    "openai-image": {
      "command": "node",
      "args": ["D:/project/fully_async/mcp-openai-image/dist/index.js"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",

        "IMAGE_OUTPUT_DIR": "D:/project/fully_async/mcp-openai-image/generated_images"
      }
    }
  }
}

Usage Examples

Architecture diagram

generate_image(
  prompt = "AWS microservices architecture diagram: React SPA → API Gateway → 3 Lambda functions (auth, orders, products) → DynamoDB + S3. Clean white background, blue/orange color scheme, connected with arrows, professional tech style.",
  size = "1536x1024",
  quality = "high"
)

System design

generate_image(
  prompt = "System design diagram for a real-time chat application: WebSocket servers, Redis pub/sub, message queue (Kafka), PostgreSQL, CDN for media. Dark background, neon connectors, modern style.",
  size = "1536x1024",
  quality = "medium"
)

Supported Parameters

Parameter	Options	Default
size	1024x1024, 1024x1536, 1536x1024, auto	1024x1024
quality	low, medium, high, auto	medium
n	1–4	1
output_format	png, jpeg, webp	png
save_to_disk	true / false	true
output_dir	any path	./generated_images