Models

Models used for image generation feature of the Lillo AI Framework on Telegram.

DALL-E 3

Configuration

const response = await openai.images.generate({
  model: 'dall-e-3',
  prompt: combinedPrompt,
  n: 1,
  size: '1024x1024',
  quality: 'standard',
  response_format: 'url'
});

Features

• High-quality image generation

• 1024x1024 resolution

• Standard quality setting

• URL response format

• Single image per request

Style Enhancement

• Template-based prompt enhancement

• Style consistency

• Quality optimization

• Content guidelines

Error Handling

function getDalleErrorMessage(error: any): string {
  if (error?.type === 'invalid_request_error') {
    if (error.code === 'content_policy_violation') {
      return '❌ Your prompt was rejected due to content policy violation...';
    }
    if (error.message?.toLowerCase().includes('rate limit')) {
      return '⏳ Rate limit exceeded...';
    }
    if (error.message?.toLowerCase().includes('capacity')) {
      return '🔄 The service is currently at capacity...';
    }
  }
  // ... other error types
}

GPT-4o-mini (Caption Generation)

Configuration

const completion = await openai.chat.completions.create({
  messages: [{ 
    role: "system", 
    content: await getCombinedPrompt('img', prompt, agent)
  }],
  model: "gpt-4o-mini",
  temperature: 0.7,
  max_tokens: 150
});

Features

• Intelligent caption generation

• Context-aware responses

• HTML formatting support

• Temperature control

• Token limit optimization

Prompt Enhancement

• System prompt integration

• Agent personality adaptation

• Style consistency

• Content guidelines

Error Handling

• Fallback to original prompt

• Error logging

• Response validation

• Format verification

Integration

Parallel Processing

const [imageUrl, caption] = await Promise.all([
  generateImage(imagePrompt),
  generateImageCaption(imagePrompt, agent)
]);

Response Handling

// Format caption with HTML
const caption = completion.choices[0].message.content || prompt;
return caption
  .replace(/\*([^*]+)\*/g, '<i>$1</i>')
  .trim();

Buffer Management

// Download and buffer image
const imageResponse = await fetch(imageUrl);
const imageBuffer = Buffer.from(await imageResponse.arrayBuffer());

Performance Considerations

Rate Limiting

• DALL-E API limits

• GPT-4o-mini throughput

• Function cooldowns

• User quotas

Optimization

• Parallel requests

• Response caching

• Buffer management

• Error recovery

Monitoring

• Request tracking

• Error logging

• Performance metrics

• Usage analytics

Related Documentation

• Overview - Feature overview

• Commands - Usage and examples