> ## Documentation Index
> Fetch the complete documentation index at: https://docs.modelslab.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Installation Guide

> Step-by-step guide to install and configure the ModelsLab MCP Server with Claude Desktop.

## Prerequisites

Before installing the ModelsLab MCP Server, ensure you have:

* **Python 3.7+** installed on your system
* **Claude Desktop** application installed
* **ModelsLab account** with API access
* **Git** for cloning the repository

## Step 1: Clone the Repository

First, clone the ModelsLab MCP Server repository to your local machine:

```bash bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
git clone https://github.com/ModelsLab/modelslab-mcp.git
```

Navigate to the cloned directory:

```bash bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
cd modelslab-mcp
```

## Step 2: Get Your API Key

1. Visit the [ModelsLab Portal](https://modelslab.com/dashboard/api-keys)
2. Log in to your account
3. Navigate to the API Keys section
4. Generate a new API key or copy your existing one
5. Keep this key secure - you'll need it for configuration

## Step 3: Create Configuration File

Create the Claude Desktop configuration file in the appropriate location:

### For macOS:

```bash bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
cd ~/Library/Application\ Support/Claude/

touch claude_desktop_config.json
```

### For Windows:

```bash bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
cd %APPDATA%\Claude\

type nul > claude_desktop_config.json
```

## Step 4: Configure the MCP Server

Add the following configuration to your `claude_desktop_config.json` file:

```json claude_desktop_config.json theme={"theme":{"light":"github-light","dark":"github-dark"}}
{
  "mcpServers": {
    "ModelsLab-MCP": {
      "command": "python",
      "args": [
        "/path/to/modelslab-mcp"
      ],
      "env": {
        "IMAGE_STORAGE_DIRECTORY": "/path/to/your/images",
        "MODELSLAB_API_KEY": "your_api_key_here"
      }
    }
  }
}
```

### Configuration Parameters:

| Parameter                 | Description                    | Example                                |
| ------------------------- | ------------------------------ | -------------------------------------- |
| `command`                 | Python executable path         | `python` or `python3`                  |
| `args`                    | Path to the cloned repository  | `"/Users/username/modelslab-mcp"`      |
| `IMAGE_STORAGE_DIRECTORY` | Where to save generated images | `"/Users/username/Pictures/ModelsLab"` |
| `MODELSLAB_API_KEY`       | Your ModelsLab API key         | `"ml-abc123..."`                       |

## Step 5: Update Paths

Replace the placeholder paths with your actual paths:

### Example for macOS:

```json claude_desktop_config.json theme={"theme":{"light":"github-light","dark":"github-dark"}}
{
  "mcpServers": {
    "ModelsLab-MCP": {
      "command": "/usr/bin/python3",
      "args": [
        "/Users/johndoe/modelslab-mcp"
      ],
      "env": {
        "IMAGE_STORAGE_DIRECTORY": "/Users/johndoe/Pictures/ModelsLab",
        "MODELSLAB_API_KEY": "ml-your-actual-api-key"
      }
    }
  }
}
```

### Example for Windows:

```json claude_desktop_config.json theme={"theme":{"light":"github-light","dark":"github-dark"}}
{
  "mcpServers": {
    "ModelsLab-MCP": {
      "command": "C:\\Python39\\python.exe",
      "args": [
        "C:\\Users\\JohnDoe\\modelslab-mcp"
      ],
      "env": {
        "IMAGE_STORAGE_DIRECTORY": "C:\\Users\\JohnDoe\\Pictures\\ModelsLab",
        "MODELSLAB_API_KEY": "ml-your-actual-api-key"
      }
    }
  }
}
```

## Step 6: Restart Claude Desktop

1. Completely quit Claude Desktop (not just minimize)
2. Reopen Claude Desktop
3. The MCP server should now be available

## Verification

To verify the installation worked:

1. Open a new conversation in Claude Desktop
2. Try generating an image with a command like:
   ```
   Generate an image of a cat sitting on a tree
   ```
3. If successful, you should see the generated image and it will be saved to your specified directory