Comprehensive comparison for Model Serving technology in AI applications

import express from 'express';
import axios from 'axios';
import rateLimit from 'express-rate-limit';
import { body, validationResult } from 'express-validator';

const app = express();
app.use(express.json());

// Configuration for LM Studio local server
const LM_STUDIO_CONFIG = {
  baseURL: process.env.LM_STUDIO_URL || 'http://localhost:1234/v1',
  timeout: 60000,
  maxRetries: 3
};

// Rate limiting to prevent abuse
const limiter = rateLimit({
  windowMs: 15 * 60 * 1000,
  max: 100,
  message: 'Too many requests from this IP'
});

app.use('/api/', limiter);

// Helper function to call LM Studio with retry logic
async function callLMStudio(messages, retries = 0) {
  try {
    const response = await axios.post(
      `${LM_STUDIO_CONFIG.baseURL}/chat/completions`,
      {
        model: 'local-model',
        messages: messages,
        temperature: 0.7,
        max_tokens: 500,
        stream: false
      },
      { timeout: LM_STUDIO_CONFIG.timeout }
    );
    return response.data;
  } catch (error) {
    if (retries < LM_STUDIO_CONFIG.maxRetries && error.code === 'ECONNABORTED') {
      console.log(`Retry attempt ${retries + 1}`);
      await new Promise(resolve => setTimeout(resolve, 1000 * (retries + 1)));
      return callLMStudio(messages, retries + 1);
    }
    throw error;
  }
}

// Product description generation endpoint
app.post(
  '/api/generate-product-description',
  [
    body('productName').trim().isLength({ min: 1, max: 200 }).escape(),
    body('features').isArray({ min: 1, max: 10 }),
    body('category').trim().isLength({ min: 1, max: 100 }).escape(),
    body('targetAudience').optional().trim().isLength({ max: 200 }).escape()
  ],
  async (req, res) => {
    // Validate input
    const errors = validationResult(req);
    if (!errors.isEmpty()) {
      return res.status(400).json({ error: 'Invalid input', details: errors.array() });
    }

    const { productName, features, category, targetAudience } = req.body;

    try {
      // Construct prompt for LM Studio
      const systemPrompt = 'You are a professional product copywriter. Generate compelling, SEO-friendly product descriptions.';
      const userPrompt = `Product: ${productName}\nCategory: ${category}\nFeatures: ${features.join(', ')}${targetAudience ? `\nTarget Audience: ${targetAudience}` : ''}\n\nGenerate a professional product description (100-150 words).`;

      const messages = [
        { role: 'system', content: systemPrompt },
        { role: 'user', content: userPrompt }
      ];

      // Call LM Studio
      const completion = await callLMStudio(messages);

      if (!completion.choices || completion.choices.length === 0) {
        throw new Error('No response from model');
      }

      const description = completion.choices[0].message.content.trim();

      // Return successful response
      res.json({
        success: true,
        description: description,
        metadata: {
          model: completion.model,
          tokensUsed: completion.usage?.total_tokens || 0,
          timestamp: new Date().toISOString()
        }
      });
    } catch (error) {
      console.error('Error generating description:', error.message);
      
      // Handle specific error types
      if (error.code === 'ECONNREFUSED') {
        return res.status(503).json({ error: 'AI service unavailable. Please try again later.' });
      }
      
      if (error.code === 'ECONNABORTED') {
        return res.status(504).json({ error: 'Request timeout. Please try again.' });
      }

      res.status(500).json({ error: 'Failed to generate description', message: error.message });
    }
  }
);

// Health check endpoint
app.get('/api/health', async (req, res) => {
  try {
    await axios.get(`${LM_STUDIO_CONFIG.baseURL}/models`, { timeout: 5000 });
    res.json({ status: 'healthy', lmStudio: 'connected' });
  } catch (error) {
    res.status(503).json({ status: 'unhealthy', lmStudio: 'disconnected' });
  }
});

const PORT = process.env.PORT || 3000;
app.listen(PORT, () => {
  console.log(`Server running on port ${PORT}`);
  console.log(`LM Studio endpoint: ${LM_STUDIO_CONFIG.baseURL}`);
});