Content
# SD WebUI MCP Server
An MCP (Model Context Protocol) server that connects Claude Desktop to your local Stable Diffusion WebUI installation, enabling AI-assisted image generation directly from your conversations.
## 🔋 Overview
This project allows you to utilize your local Stable Diffusion installation through Claude Desktop, giving you the ability to:
* Generate images from text descriptions
* Edit existing images with text prompts
* List and switch between different Stable Diffusion models
* Access advanced image generation parameters
* View image thumbnails directly in Claude responses
All while staying within the Claude Desktop interface, without switching between applications.
## 🎯 Features
* **Text-to-Image Generation**: Create images from text prompts
* **Image-to-Image Editing**: Modify existing images using text guidance
* **Model Management**: List and switch between your installed models
* **Sampler Control**: Access all WebUI samplers
* **Connection Status**: Verify connectivity with your local Stable Diffusion
* **High Performance**: Uses your local GPU resources for image generation
* **Privacy-Focused**: All processing happens on your machine
* **Base64 Thumbnails**: See images directly in Claude responses
* **SDXL Optimized**: Default settings tuned for SDXL's 1024x1024 resolution
## 📦 Prerequisites
* Node.js v16.0.0 or higher
* Stable Diffusion WebUI with API enabled
* Claude Desktop
* Basic familiarity with command line interfaces
## 🚀 Installation
### 1. Clone this repository
```bash
git clone https://github.com/boxi-rgb/sd-webui-mcp.git
cd sd-webui-mcp
```
### 2. Install dependencies
```bash
npm install
```
### 3. Build the project
```bash
npm run build
```
### 4. Configure Stable Diffusion WebUI
Make sure your Stable Diffusion WebUI is running with the API enabled.
#### For Windows:
Edit your `webui-user.bat` file and add `--api` to the COMMANDLINE_ARGS:
```
@echo off
set COMMANDLINE_ARGS=--api
call webui.bat
```
#### For Linux/macOS:
Edit your `webui-user.sh` file and add `--api` to the COMMANDLINE_ARGS:
```bash
export COMMANDLINE_ARGS="--api"
./webui.sh
```
### 5. Configure Claude Desktop
1. Open Claude Desktop
2. Go to Settings
* Windows: `%APPDATA%\Claude\claude_desktop_config.json`
* macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
3. Add the following to the `claude_desktop_config.json` file (create it if it doesn't exist):
```json
{
"mcpServers": {
"stable-diffusion": {
"command": "node",
"args": [
"/FULL/PATH/TO/sd-webui-mcp/dist/index.js"
]
}
}
}
```
**Important**: Replace `/FULL/PATH/TO/` with the actual full path to your project directory. For Windows users, use double backslashes (`\\`) or forward slashes (`/`) in your path.
4. Save the file and restart Claude Desktop
## ⚙️ Usage
### Step 1: Verify installation
After starting Claude Desktop, you should see a tools icon (🔧) in the bottom right corner of the chat interface. Click on it to see if "stable-diffusion" appears in the list of available tools.
### Step 2: Check connection
Type the following message to Claude:
```
Can you check if Stable Diffusion is connected properly?
```
Claude should use the `check-status` tool to verify the connection with your local Stable Diffusion WebUI.
### Step 3: Generate your first image
Type something like:
```
Please create an image of a futuristic city with flying cars.
```
Claude will use the `generate-image` tool to send your request to Stable Diffusion and display the resulting image.
### Advanced Usage
The MCP server provides several tools that Claude can use:
#### 1. Image Generation
```
Generate an image of a cat wearing a space suit on Mars.
```
You can add more detailed parameters:
```
Generate an image of a mountain landscape with the following parameters:
- Width: 1024
- Height: 1024
- Steps: 30
- CFG Scale: 8
- Negative prompt: blurry, low quality
```
#### 2. Model Switching
First, check available models:
```
What Stable Diffusion models are available on my system?
```
Then switch models:
```
Please switch to the AnythingV5 model.
```
#### 3. Editing Existing Images
Save an image to your system, then:
```
Edit the image at C:/path/to/image.png to add flowers in the background.
```
#### 4. List Sampling Methods
```
What samplers are available in my Stable Diffusion setup?
```
## 🔧 Configuration
You can customize the behavior of the MCP server by modifying `src/config.ts`:
* `SD_API_URL`: The URL of your Stable Diffusion WebUI API (default: http://127.0.0.1:7860)
* `OUTPUT_DIR`: Directory where generated images will be saved (default: C:\\SD_Output)
* `DEFAULT_PARAMS`: Default image generation parameters optimized for SDXL (1024x1024 resolution)
After modifying, rebuild the project with `npm run build`.
## 🔍 Troubleshooting
### Common Issues
#### MCP server not found in Claude Desktop
* Verify the path in `claude_desktop_config.json` is correct
* Ensure you've included the full absolute path to the `dist/index.js` file
* Check if the Node.js path is correct
#### Cannot connect to Stable Diffusion WebUI
* Make sure Stable Diffusion WebUI is running
* Verify that it was started with the `--api` flag
* Check if the default port (7860) is being used or if you need to specify a custom port in `config.ts`
#### Images not being generated
* Check the output directory exists and has write permissions
* Look for error messages in the Claude Desktop console
* Try running the `check-status` tool to verify connectivity
### Debugging
If you're experiencing issues, you can enable more verbose logging by:
1. Running the MCP server directly from the command line to see error messages:
```bash
node dist/index.js
```
2. You can also use the MCP Inspector for debugging:
```bash
npx @modelcontextprotocol/inspector node /path/to/sd-webui-mcp/dist/index.js
```
## 📝 License
MIT License
## ⚠️ Disclaimer
This project is not officially affiliated with Anthropic (creators of Claude) or Stable Diffusion. Use at your own risk.
Connection Info
You Might Also Like
markitdown
Python tool for converting files and office documents to Markdown.
OpenAI Whisper
OpenAI Whisper MCP Server - 基于本地 Whisper CLI 的离线语音识别与翻译,无需 API Key,支持...
oh-my-opencode
Background agents · Curated agents like oracle, librarians, frontend...
servo-fetch
A self-contained browser engine that fetches, renders, and extracts web...
deep-code-reasoning-mcp
A Model Context Protocol (MCP) server that provides advanced code analysis...
memoriki
Memoriki - LLM Wiki + MemPalace. Personal knowledge base with real memory.