Medium Models (8.1B-20B Parameters)

Our medium-sized models offer enhanced capabilities and specialized features. With 8.1B to 20B parameters, these models provide superior performance for specific use cases while maintaining reasonable resource requirements.

Available Models

Qwen-14B

• Parameters: 14B
• Context Window: 8192 tokens
• Provider: Alibaba Cloud
• License: Apache 2.0
• Key Features:
  • Strong multilingual capabilities
  • Advanced reasoning
  • Knowledge integration
  • Technical understanding

DeepSeek Math 14B

• Parameters: 14B
• Context Window: 4096 tokens
• Provider: DeepSeek AI
• License: Apache 2.0
• Key Features:
  • Mathematical specialization
  • Step-by-step solutions
  • Formula understanding
  • Scientific computation

Configuration & Setup

OpenAI SDK Setup

import OpenAI from 'openai';

const openai = new OpenAI({
  apiKey: 'your-neuredge-key',
  baseURL: 'https://api.neuredge.dev/v1/'
});

Native SDK Setup

import { Neuredge } from '@neuredge/sdk';

const neuredge = new Neuredge({
  apiKey: 'your-api-key'
});

Real-World Applications & Examples

1. Advanced Mathematical Problem Solving

Step-by-Step Math Solutions

const response = await openai.chat.completions.create({
  model: '@cf/deepseek-math-7b-instruct',
  messages: [
    {
      role: 'system',
      content: 'You are a mathematics tutor. Provide clear, step-by-step solutions.'
    },
    {
      role: 'user',
      content: \`Solve this calculus problem:
      Find the volume of the solid obtained by rotating the region bounded by 
      y = x², y = 2x, and x = 0 about the x-axis.\`
    }
  ],
  temperature: 0.3,
  max_tokens: 800
});

Use Cases:

• Math education
• Engineering calculations
• Scientific computing
• Research analysis
• Statistical modeling

2. Multilingual Content Creation

Global Marketing Content

const response = await openai.chat.completions.create({
  model: '@cf/qwen/qwen1.5-14b-chat-awq',
  messages: [
    {
      role: 'system',
      content: 'You are a multilingual marketing expert. Create culturally appropriate content.'
    },
    {
      role: 'user',
      content: \`Create product descriptions for a luxury watch in:
      1. English (US market)
      2. Chinese (mainland China market)
      3. Japanese (Japan market)
      Focus on cultural preferences and market-specific value propositions.\`
    }
  ],
  temperature: 0.7,
  max_tokens: 1000
});

Use Cases:

• International marketing
• Localization
• Cultural adaptation
• Global communications
• Multilingual documentation

3. Technical Documentation & Analysis

API Documentation Generator

const response = await openai.chat.completions.create({
  model: '@cf/qwen/qwen1.5-14b-chat-awq',
  messages: [
    {
      role: 'system',
      content: 'Generate comprehensive API documentation with examples and best practices.'
    },
    {
      role: 'user',
      content: \`Create documentation for this GraphQL API schema:
      ${schemaDefinition}
      Include:
      - Field descriptions
      - Usage examples
      - Common queries
      - Error handling\`
    }
  ],
  temperature: 0.4,
  max_tokens: 1500
});

Use Cases:

• Technical writing
• API documentation
• System specifications
• Architecture documents
• Implementation guides

4. Scientific Research Assistance

Research Analysis

const response = await openai.chat.completions.create({
  model: '@cf/deepseek-math-7b-instruct',
  messages: [
    {
      role: 'system',
      content: 'You are a research assistant analyzing scientific papers and data.'
    },
    {
      role: 'user',
      content: \`Analyze this dataset and research hypothesis:
      ${experimentData}
      Provide:
      1. Statistical analysis
      2. Correlation findings
      3. Potential implications
      4. Suggested next steps\`
    }
  ],
  temperature: 0.4,
  max_tokens: 1000
});

Use Cases:

• Data analysis
• Research synthesis
• Experimental design
• Statistical review
• Methodology planning

Integration Examples

Flask Research Application

from flask import Flask, request, jsonify
from openai import OpenAI

app = Flask(__name__)
client = OpenAI(
    api_key='your-neuredge-key',
    base_url='https://api.neuredge.dev/v1/'
)

@app.route('/analyze-research', methods=['POST'])
def analyze_research():
    try:
        data = request.json
        response = client.chat.completions.create(
            model='@cf/deepseek-math-7b-instruct',
            messages=[
                {
                    'role': 'system',
                    'content': 'Analyze research data and provide insights.'
                },
                {
                    'role': 'user',
                    'content': f"Analyze this research data:\n{data['content']}"
                }
            ],
            temperature=0.4
        )
        
        return jsonify({
            'analysis': response.choices[0].message.content,
            'usage': response.usage
        })
    except Exception as e:
        return jsonify({'error': str(e)}), 500

Next.js Multilingual Content Generator

// pages/api/generate-content.js
import { OpenAIStream, StreamingTextResponse } from 'ai';
import OpenAI from 'openai';

const openai = new OpenAI({
  apiKey: process.env.NEUREDGE_API_KEY,
  baseURL: 'https://api.neuredge.dev/v1/'
});

export default async function handler(req, res) {
  const { content, languages } = await req.json();

  const response = await openai.chat.completions.create({
    model: '@cf/qwen/qwen1.5-14b-chat-awq',
    messages: [
      {
        role: 'system',
        content: 'Translate and adapt content for different markets.'
      },
      {
        role: 'user',
        content: `Content: ${content}\nTarget languages: ${languages.join(', ')}`
      }
    ],
    stream: true
  });

  const stream = OpenAIStream(response);
  return new StreamingTextResponse(stream);
}

Best Practices

1. Task-Specific Optimization

   • Use DeepSeek Math for mathematical/scientific tasks
   • Choose Qwen for multilingual/technical content
   • Adjust temperature based on task precision needs
   • Implement proper context handling

2. Performance Management

   • Cache complex computations
   • Use streaming for long outputs
   • Implement timeout handling
   • Monitor resource usage

3. Quality Control

   • Validate mathematical results
   • Verify translations
   • Review technical accuracy
   • Implement feedback loops

Token Management

Plan	Monthly Token Quota
Free Tier	150K tokens
$29 Plan	1.5M tokens
$49 Plan	2.25M tokens

When to Use

✅ Ideal For:

• Mathematical computations
• Scientific research
• Technical documentation
• Multilingual content
• Complex analysis

❌ Consider Alternatives When:

• Basic text generation needed
• Speed is critical
• Limited budget
• Simple queries suffice

Getting Started

To begin using our medium models:

• Quick Start Guide
• JavaScript SDK
• Python SDK