muapi MCP Server

Connect any MCP-compatible AI assistant to muapi.ai — access 100+ generative AI models (image, video, audio, enhance, edit) directly from Claude, Cursor, Windsurf, and more.

Quick Start (Hosted — Recommended)

The fastest way: use the hosted MCP server at https://api.muapi.ai/mcp. No installation required — just add it to your client.

Get your API key at muapi.ai/dashboard/keys

Claude Code

claude mcp add --transport http muapi \
  https://api.muapi.ai/mcp \
  --header "Authorization: Bearer YOUR_MUAPI_KEY"

Cursor

Open Cmd+Shift+P → Open MCP settings and add to mcp.json:

{
  "mcpServers": {
    "muapi": {
      "url": "https://api.muapi.ai/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_MUAPI_KEY"
      }
    }
  }
}

Windsurf

Open Settings → MCP and add:

{
  "mcpServers": {
    "muapi": {
      "serverUrl": "https://api.muapi.ai/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_MUAPI_KEY"
      }
    }
  }
}

Claude Desktop / Other clients

Add to your MCP client config:

{
  "mcpServers": {
    "muapi": {
      "url": "https://api.muapi.ai/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_MUAPI_KEY"
      }
    }
  }
}

Self-Hosted Setup

If you want to run your own MCP server instance:

1. Clone the repo

git clone https://github.com/SamurAIGPT/muapi-mcp-server.git
cd muapi-mcp-server

2. Install dependencies

pip install fastapi uvicorn requests pydantic

3. Set your API key

export MUAPIAPP_API_KEY=your_muapi_key_here

4. Start the server

python mcp_server.py

The server starts at http://localhost:8000.

5. Configure your client

Point your MCP client to your local server:

{
  "mcpServers": {
    "muapi": {
      "url": "http://localhost:8000",
      "headers": {
        "x-api-key": "YOUR_MUAPI_KEY"
      }
    }
  }
}

Environment Variables

Available Tools

Once connected, your AI assistant has access to these tools:

Discovery

Image Generation

Video Generation

Audio

Image Enhancement

Video Editing

Async Polling

Account & Keys

Usage Examples

Once your MCP client is connected, you can ask your AI assistant naturally:

Generate an image:

"Generate a photorealistic mountain lake at golden hour using flux-dev"

Edit an image:

"Convert this image to Studio Ghibli style" (attaches image URL)

Create a video:

"Make a 5-second cinematic video of a robot walking through a forest using kling-master"

Animate an image:

"Turn this product photo into a 5-second video" (attaches image URL)

Create music:

"Create a 30-second lo-fi hip hop track, instrumental"

Check balance:

"What's my muapi credit balance?"

Find models:

"What image generation models are available on muapi?"

How Async Generation Works

Most generation tools return a request_id immediately:

{ "request_id": "abc123", "status": "processing" }

Your assistant will automatically poll muapi_predict_result until the job completes:

{
  "request_id": "abc123",
  "status": "completed",
  "outputs": ["https://cdn.muapi.ai/..."]
}

API Reference

The self-hosted server exposes these endpoints:

Authentication

Hosted server — pass your API key in the Authorization header:

Authorization: Bearer YOUR_MUAPI_KEY

Self-hosted server — pass via x-api-key header or set MUAPIAPP_API_KEY environment variable.

Get your API key at muapi.ai/dashboard/keys.

Links

• muapi.ai — Platform
• API Documentation — Full REST API reference
• muapi CLI — Command-line interface
• Dashboard — Manage keys and credits