piapi

You run a dedicated MCP server that connects PiAPI’s API to your MCP-compatible clients, enabling media generation workflows directly from Claude or other MCP-enabled apps. This server is implemented in TypeScript, integrates securely with PiAPI, and exposes an MCP endpoint you can run locally or remotely to execute image generation actions from text prompts and other capabilities.

How to use

Set up your MCP client to connect to the server you run locally or on a remote host. Start the server with your API key available as an environment variable, then configure your MCP client to point to the server via a stdio configuration (local process) as shown in the Claude Desktop integration snippet. Once connected, you can issue Flux image generation requests from text prompts and extend to additional capabilities as they become available.

How to install

Follow these concrete steps to install and run the MCP server locally.

# Step 1: Clone the repository
 git clone https://github.com/apinetwork/piapi-mcp-server
 cd piapi-mcp-server

# Step 2: Install dependencies
 npm install

# Step 3: Create environment variables file
 echo "PIAPI_API_KEY=your_api_key_here" > .env

# Step 4: Build the project
 npm run build

# Step 5: Start the server
 npm start

Claude Desktop integration snippet

Configure Claude Desktop to run the MCP server as a local process. Create or update the configuration file so Claude can start the server with the correct runtime and environment variable. Use your absolute path to the built server entry point and your API key.

{
  "mcpServers": {
    "piapi": {
      "command": "node",
      "args": ["/absolute/path/to/piapi-mcp-server/dist/index.js"],
      "env": {
        "PIAPI_API_KEY": "your_api_key_here"
      }
    }
  }
}

Available tools

flux_image

Generate Flux images from text descriptions using PiAPI integration.

midjourney_image

(Future) Generate images using Midjourney-compatible prompts through the MCP bridge.

kling_video

(Future) Create video content via Kling integration through MCP.

luma_image_workflow

(Future) Orchestrate Luma-based image/video workflows via MCP.

udio_song

(Future) Generate AI-assisted songs using Udio or similar services through MCP.