Openrouter MCP Multimodal

🚀 OpenRouter MCP Multimodal Server

An OpenRouter-based Multimodal (MCP) server that supports text chat and image analysis functions.

✨ Features

• Text Chat: Supports dialogue interactions with multiple models integrated with OpenRouter.
• Image Analysis: Capable of processing and analyzing image content, supporting multiple image formats and sources.
• Multimodal Interaction: Supports mixed input of text and images simultaneously, providing richer application scenarios.
• Flexible Configuration: Allows custom settings for parameters such as the default model and API key to meet different development needs.

📦 Installation

1. Install via npm

npm install @openrouter/mcp-multimodal-server

2. Create a server instance

const { OpenRouterMCP } = require('@openrouter/mcp-multimodal-server');

// Initialize the MCP service
const openRouterMCP = new OpenRouterMCP({
  apiKey: 'your-api-key',
  defaultModel: 'google/palm-0.1'
});

// Start the service
async function start() {
  await openRouterMCP.listen(3000);
  console.log('OpenRouter MCP server has been started and is listening on port 3000');
}
start();

💻 Usage Examples

Basic Usage

Text Chat

const result = await openRouterMCP.chatCompletion({
  model: 'anthropic/claude-3',
  messages: [
    {
      role: 'system',
      content: 'You are a very helpful assistant.'
    },
    {
      role: 'user',
      content: 'What is the capital of France?'
    }
  ],
  temperature: 0.7
});

console.log('Answer:', result.choices[0].message.content);

Multimodal Dialogue

const result = await openRouterMCP.chatCompletion({
  model: 'meta/llama2',
  messages: [
    {
      role: 'user',
      content: [
        { type: 'text', text: 'What is in the picture?' },
        {
          type: 'image_url',
          image_url: {
            url: 'https://example.com/image.jpg'
          }
        }
      ]
    }
  ],
  temperature: 0.5
});

console.log('Answer:', result.choices[0].message.content);

📚 Documentation

Configuration Options

const openRouterMCP = new OpenRouterMCP({
  apiKey: 'your-api-key', // Required: OpenRouter API key
  defaultModel: 'google/palm-0.1', // Optional: Default model to use
  enableImageProcessing: true, // Enable image processing function
  imageProcessingConcurrency: 5 // Concurrency for image processing, default is 5
});

Supported Models

Currently, all available models on OpenRouter are supported, including but not limited to:

• Text Models:
  • google/palm-0.1
  • anthropic/claude-3
  • meta/llama2
• Multimodal Models:
  • microsoft/vision-language-imagenet-v1
  • openai/gpt-4-vision

Error Handling

try {
  const result = await openRouterMCP.chatCompletion({
    model: 'non-existent-model',
    messages: [{ role: 'user', content: 'Test error handling' }]
  });
} catch (error) {
  console.error('Error:', error.message);
}

📄 License

MIT License