Openai Image Generation

🚀 OpenAI Image Generation MCP Server

This project implements an MCP (Model Context Protocol) server that provides tools for generating and editing images using OpenAI's gpt-image-1 model via the official Python SDK.

✨ Features

This MCP server offers the following tools:

generate_image

Generate an image based on a text prompt using OpenAI's gpt-image-1 model and save it.

• Input Schema:

{
  "type": "object",
  "properties": {
    "prompt": { "type": "string", "description": "Text description of the desired image." },
    "model": { "type": "string", "default": "gpt-image-1", "description": "The model to use (currently 'gpt-image-1')." },
    "n": { "type": ["integer", "null"], "description": "Number of images to generate.", "minimum": 1 },
    "size": { "type": "string", "enum": ["512x512", "1024x1024"] },
    "response_format": { "type": "string", "enum": ["url", "b64_json"], "default": "url" }
  },
  "required": ["prompt"]
}

• Output: {"status": "success", "output": "Path or URL of the generated image"} or an error dictionary.

edit_image

Edit an image based on a prompt and an input image using OpenAI's gpt-image-1 model and save the result.

• Input Schema:

{
  "type": "object",
  "properties": {
    "prompt": { "type": "string", "description": "Prompt for editing the image." },
    "image_path": { "type": "string", "description": "Path to the image to be edited." },
    "response_format": { "type": "string", "enum": ["url", "b64_json"], "default": "url" }
  },
  "required": ["prompt", "image_path"]
}

• Output: {"status": "success", "output": "Path or URL of the edited image"} or an error dictionary.

📦 Installation

Prerequisites

• Python (version 3.8 or higher is recommended).
• pip (Python package installer).
• An OpenAI API key (set directly in the script or via the OPENAI_API_KEY environment variable - setting via the environment variable is crucial for security).
• An MCP client environment capable of managing and starting the MCP server (e.g., the environment used by Cline).

Steps

1. Clone the repository:

git clone https://github.com/IncomeStreamSurfer/chatgpt-native-image-gen-mcp.git
cd chatgpt-native-image-gen-mcp

2. Set up a virtual environment (recommended):

python -m venv venv
source venv/bin/activate  # Use `venv\Scripts\activate` on Windows

3. Install dependencies:

pip install -r requirements.txt

4. (Optional but recommended) Set environment variables: Set the OPENAI_API_KEY environment variable instead of hard - coding the API key in the script. The method of setting this environment variable varies depending on your operating system.

📚 Configuration (for Cline MCP Client)

To make this server available to your AI assistant (e.g., Cline), add its configuration to the MCP settings file (e.g., cline_mcp_settings.json).

Add the following entry under the mcpServers object in the settings file:

{
  "mcpServers": {
    // ... Other server configurations ...

    "openai-image-gen-mcp": {
      "autoApprove": [
        "generate_image",
        "edit_image"
      ],
      "disabled": false,
      "timeout": 180, // Increase the timeout as image generation may take a long time
      "command": "python", // Or the path to the Python executable if not in PATH
      "args": [
        // Key: Replace this path with the actual absolute path of your cloned repository
        // to the `openai_image_mcp.py` file on your machine
        "C:/path/to/your/cloned/repo/chatgpt-native-image-gen-mcp/openai_image_mcp.py"
      ],
      "env": {
        // If you set the API key via an environment variable:
        // "OPENAI_API_KEY": "YOUR_API_KEY_HERE"
      },
      "transportType": "stdio"
    }

    // ... Other server configurations ...
  }
}

⚠️ Important Note

Replace C:/path/to/your/cloned/repo/ with the actual absolute path of the cloned repository on your machine. Ensure that the path separators are appropriate for your operating system (e.g., use backslashes \ on Windows). If you set the API key via an environment variable, you can remove it from the script and potentially add it to the env section here if your MCP client supports it.

🚀 Quick Start

Normally, you don't need to manually run the server. The MCP client (e.g., Cline) will automatically start the server using the command and args in the configuration file when the first tool is called.