Image Generation

Our image generation capabilities provide state-of-the-art models for creating high-quality images from text descriptions. We offer both standard and fast generation options to suit different use cases.

Available Models

Model	Speed	Quality	Resolution	Use Case
SDXL Base	Standard	Best	1024x1024	Production quality
SDXL Lightning	2x Faster	Great	1024x1024	Rapid prototyping

Model Selection Guide

• SDXL Base (Standard)

  • Provider: Stability AI
  • Steps: 20 (configurable)
  • Best for: High-quality production images
  • Guidance scale: 7.5 (configurable)

• SDXL Lightning (Fast)

  • Provider: ByteDance
  • Steps: 10 (optimized)
  • Best for: Quick iterations, prototyping
  • Guidance scale: 7.5 (optimized)

Basic Usage

Native SDK

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

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

const image = await neuredge.generateImage({
  model: '@cf/stabilityai/stable-diffusion-xl-base-1.0',
  prompt: 'A serene lake at sunset with mountains in the background',
  width: 1024,
  height: 1024,
  num_steps: 20,
  guidance: 7.5
});

REST API

const response = await fetch('https://api.neuredge.dev/v1/images/generations', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${apiKey}`,
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    model: '@cf/stabilityai/stable-diffusion-xl-base-1.0',
    prompt: 'A cyberpunk cityscape at night with neon signs',
    width: 1024,
    height: 1024,
    num_steps: 20
  })
});

const image = await response.blob();

Real-World Applications & Examples

1. E-commerce Product Visualization

async function generateProductVariant(product, color) {
  const response = await neuredge.generateImage({
    model: '@cf/stabilityai/stable-diffusion-xl-base-1.0',
    prompt: \`${product.name} in ${color} color, professional product photography, 
    white background, high resolution, detailed texture\`,
    width: 1024,
    height: 1024,
    num_steps: 20,
    guidance: 7.5
  });
  
  return response;
}

Use Cases:

• Product mockups
• Color variations
• Marketing materials
• Catalog imagery
• Design prototypes

2. Real Estate Visualization

async function generateInteriorDesign(specs) {
  const response = await neuredge.generateImage({
    model: '@cf/stabilityai/stable-diffusion-xl-base-1.0',
    prompt: \`${specs.style} style ${specs.room} with ${specs.features.join(', ')}, 
    interior design, photorealistic, natural lighting\`,
    width: 1024,
    height: 1024,
    num_steps: 20,
    guidance: 8.0 // Slightly higher for more photorealism
  });
  
  return response;
}

Use Cases:

• Interior design concepts
• Property staging
• Renovation previews
• Architectural visualization
• Design alternatives

3. Marketing Content Creation

async function generateSocialMedia(campaign) {
  // For quick iterations during brainstorming
  const draft = await neuredge.generateImage({
    model: '@cf/bytedance/stable-diffusion-xl-lightning',
    prompt: campaign.concept,
    width: 1024,
    height: 1024,
    num_steps: 10
  });

  // For final production version
  const final = await neuredge.generateImage({
    model: '@cf/stabilityai/stable-diffusion-xl-base-1.0',
    prompt: campaign.finalPrompt,
    width: 1024,
    height: 1024,
    num_steps: 20,
    guidance: 7.5
  });

  return { draft, final };
}

Use Cases:

• Social media content
• Ad creatives
• Marketing materials
• Campaign assets
• Brand imagery

Integration Examples

Express.js Image Generation API

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

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

const neuredge = new Neuredge({
  apiKey: process.env.NEUREDGE_API_KEY
});

app.post('/generate-image', async (req, res) => {
  try {
    const { prompt, style = 'standard' } = req.body;
    
    const model = style === 'fast' 
      ? '@cf/bytedance/stable-diffusion-xl-lightning'
      : '@cf/stabilityai/stable-diffusion-xl-base-1.0';

    const response = await neuredge.generateImage({
      model,
      prompt,
      width: 1024,
      height: 1024,
      num_steps: style === 'fast' ? 10 : 20
    });

    // Convert to Base64 for JSON response
    const base64 = await response.arrayBuffer()
      .then(buffer => Buffer.from(buffer).toString('base64'));

    res.json({
      image: \`data:image/png;base64,\${base64}\`,
      model
    });
  } catch (error) {
    res.status(500).json({ error: error.message });
  }
});

Next.js Image Generation

// pages/api/generate.js
import { Neuredge } from '@neuredge/sdk';

const neuredge = new Neuredge({
  apiKey: process.env.NEUREDGE_API_KEY
});

export default async function handler(req, res) {
  if (req.method !== 'POST') {
    return res.status(405).json({ message: 'Method not allowed' });
  }

  try {
    const { prompt, style = 'standard' } = req.body;

    const response = await neuredge.generateImage({
      model: style === 'fast'
        ? '@cf/bytedance/stable-diffusion-xl-lightning'
        : '@cf/stabilityai/stable-diffusion-xl-base-1.0',
      prompt,
      width: 1024,
      height: 1024,
      num_steps: style === 'fast' ? 10 : 20
    });

    // Set proper content type for image response
    res.setHeader('Content-Type', 'image/png');
    res.send(await response.arrayBuffer());
  } catch (error) {
    res.status(500).json({ error: error.message });
  }
}

Best Practices

1. Prompt Engineering

   • Be specific and detailed
   • Include desired style
   • Specify important details
   • Use descriptive adjectives

2. Model Selection

   • Use Lightning for drafts
   • SDXL Base for final images
   • Consider use case requirements
   • Balance speed vs quality

3. Performance Optimization

   • Cache generated images
   • Use appropriate dimensions
   • Optimize step count
   • Implement retry logic

4. Error Handling

   • Handle failed generations
   • Provide fallback options
   • Monitor error rates
   • Log issues for analysis

Quotas and Limits

Plan	Monthly Generation Quota
Free Tier	50 images
$29 Plan	500 images
$49 Plan	750 images

Resolution Limits

• Maximum: 1024x1024 pixels
• Minimum: 512x512 pixels
• Aspect ratio: 1:1 (square)

When to Use Each Model

SDXL Base (Standard)

✅ Ideal For:

• Production content
• Marketing materials
• Client deliverables
• Brand assets
• Professional use

SDXL Lightning (Fast)

✅ Ideal For:

• Rapid prototyping
• Design iterations
• Concept exploration
• Quick drafts
• Testing ideas

Getting Started

1. Install the SDK:

npm install @neuredge/sdk
# or
pip install neuredge-sdk

2. Initialize the client:

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

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

Learn more:

• Quick Start Guide
• JavaScript SDK
• Python SDK