Pollinations Multimodal

🚀 MCPollinations MCP Server

MCPollinations MCP Server is a powerful tool that offers a range of capabilities for image generation, text response, audio generation, and model management.

🚀 Quick Start

MCPollinations MCP Server is a versatile tool with multiple functions. It can generate high - quality images, handle text instructions through the MCP protocol, generate audio content, and manage models flexibly.

✨ Features

1. Image Generation: Generate high - quality images using different models (default is 'flux'). It supports customizing image size, random seed, and content filtering.
2. Text Response: Interact with client applications (such as Claude Desktop) via the MCP protocol to process text instructions.
3. Audio Generation: Provide a variety of voice options for generating voice content.
4. Model Management: Support flexible model selection and image format settings.

📦 Installation

Installation Steps

1. Install Dependencies

   npm install @pinkpixel/mcpollinations
   

2. Import Modules

   import { generateImageUrl, generateImage, respondText, respondAudio, listTextModels, listImageModels, listAudioVoices } from '@pinkpixel/mcpollinations';
   

3. Start the Server

   • Run in the terminal:

     npm start
     

   • Default port: 8080
   • Access address: http://localhost:8080

Environment Requirements

• Operating System: Supports Linux, macOS, and Windows.
• Node.js Version: Node.js 14.19.1 or higher is required.

📚 Documentation

Image Generation Options

const options = {
  // Model selection (default is 'flux')
  model: "flux",

  // Image size
  width: 1024,
  height: 1024,

  // Generation parameters
  seed: 12345,  // Specific seed value for ensuring reproducibility (default is random)
  enhance: true,  // Use LLM to enhance the prompt (default is enabled)
  safe: false,  // Content filtering switch (default is disabled)

  // File saving options
  saveToFile: true,  // Set to false to skip file saving
  outputPath: "/path/to/save/directory",  // Custom save path
  fileName: "my_custom_name",  // Without extension
  format: "png"  // Supported formats: png, jpeg, jpg, webp
};

Output Path Explanation

• Default Location: The current working directory under the server running directory.
• When Claude Desktop starts automatically: Saved in Claude's application data folder (usually under the user's home directory).

Duplication Prevention Mechanism

To ensure the uniqueness of generated images and avoid overwriting, the MCP server takes the following measures:

1. Default File Name:
   • A purified version of the prompt (first 20 characters)
   • Timestamp
   • Random suffix
2. Custom File Name Protection:
   • If the specified file already exists, a numeric suffix is automatically added.
   • Example: sunset.png, sunset_1.png, etc.

Base64 Data Access

Regardless of whether it is saved to a file, Base64 - encoded image data is returned, which is convenient for the following purposes:

• Embedding in web pages (<img>)
• Passing to other services or APIs
• In - memory processing
• Display in applications that support data URIs

💻 Usage Examples

Basic Usage

Integrate MCPollinations into your project by following these steps:

1. Install Dependencies

   npm install @pinkpixel/mcpollinations
   

2. Code Example

   import { generateImageUrl, generateImage, respondText, respondAudio, listTextModels, listImageModels, listAudioVoices } from '@pinkpixel/mcpollinations';
   
   // Usage example:
   const imageUrl = generateImageUrl('sunset over mountains', {
     width: 800,
     height: 600
   });
   

🔧 Troubleshooting

Common Issues

• Port Conflict: If the port 8080 is occupied when starting, try changing to another port.
• Module Not Found: Make sure the MCPollinations package is installed correctly.

Contact Support

If you need help, please visit the GitHub repository to view the documentation or submit an issue.