Image Generation with AI Chatbots

Overview

Give your AI chatbot the ability to create images. Request artwork, generate profile pictures, create diagrams, or visualize ideas - all through simple chat messages.

How It Works

Chat Request: "Generate a sunset over mountains"
    ↓
[Your VPS - OpenClaw]
    ↓
[AI improves prompt]
    ↓
[Image Generation API (DALL-E/SD)]
    ↓
Image sent back to chat

Provider Options

| Provider | Quality | Speed | Cost | |----------|---------|-------|------| | DALL-E 3 | Excellent | Fast | $0.04-0.12/image | | Midjourney | Best artistic | Medium | $10-30/month | | Stable Diffusion | Good | Fast | Free (self-hosted) | | Google Imagen | Excellent | Fast | Pay per image |

Setup Guide

DALL-E 3 (Recommended)

Easiest setup, great quality:

# Image Generation
IMAGE_GENERATION_ENABLED=true
IMAGE_PROVIDER=openai
OPENAI_API_KEY=your-openai-key

# DALL-E Settings
DALLE_MODEL=dall-e-3
DALLE_SIZE=1024x1024
DALLE_QUALITY=standard  # or 'hd'

Stable Diffusion (Self-Hosted)

Free but requires GPU or patience:

IMAGE_PROVIDER=stable-diffusion
SD_API_URL=http://localhost:7860

Run SD locally (requires GPU) or use a hosted API.

Google Imagen

IMAGE_PROVIDER=google
GOOGLE_API_KEY=your-key
IMAGEN_MODEL=imagen-3

Usage Examples

Basic Generation

You: "Generate an image of a cozy coffee shop" Bot: [Sends image] "Here's your cozy coffee shop. Would you like any changes?"

Detailed Prompts

You: "Create a futuristic cityscape at night, cyberpunk style, neon lights, rain-slicked streets" Bot: [Sends detailed cyberpunk cityscape]

AI-Enhanced Prompts

Your bot can improve prompts automatically:

ENHANCE_IMAGE_PROMPTS=true

You: "Make a cat" Bot internally enhances to: "A photorealistic orange tabby cat sitting on a windowsill, soft natural lighting, detailed fur, green eyes, cozy home setting" Bot: [Sends beautiful cat image]

Specific Styles

You: "Draw a castle in watercolor style" You: "Generate my logo in pixel art" You: "Create a portrait in oil painting style"

Bot Self-Portrait

Fun feature - let your bot generate its own avatar:

You: "Create a profile picture for yourself" Bot: "I'll create an avatar that represents me..." [Generates image]

Advanced Features

Image Variations

Request variations of an image:

You: [Sends image] "Create 3 variations of this" Bot: [Sends 3 similar but different images]

Image Editing

Edit existing images:

You: [Sends photo] "Add a santa hat to this person" Bot: [Sends edited image]

Image Analysis + Generation

Combine vision and generation:

You: [Sends room photo] "Generate this room in a minimalist style" Bot: Analyzes the room, then generates a minimalist version.

Batch Generation

You: "Generate 4 logo concepts for a tech startup called 'NexGen'" Bot: [Sends 4 different logo concepts]

Configuration Options

Quality vs Speed

# For best quality (slower, more expensive)
DALLE_QUALITY=hd
DALLE_SIZE=1024x1792

# For quick previews (faster, cheaper)
DALLE_QUALITY=standard
DALLE_SIZE=1024x1024

Size Options

DALL-E 3 sizes:

1024x1024 - Square (default)
1024x1792 - Portrait
1792x1024 - Landscape

Style Presets

IMAGE_STYLE_PRESETS="
photorealistic: Ultra realistic, professional photography, detailed
artistic: Artistic interpretation, painterly, creative
minimal: Clean, minimalist, simple shapes, modern
vintage: Retro style, film grain, muted colors
"

You: "Generate a forest, style: vintage" Bot: [Generates vintage-style forest image]

Content Filtering

# Prevent inappropriate content
IMAGE_SAFETY_FILTER=true
BLOCK_NSFW_REQUESTS=true

Use Cases

Daily Briefing Art

Generate unique artwork for daily briefings:

async function generateBriefingArt() {
  const today = new Date().toLocaleDateString('en-GB', { weekday: 'long' });
  const weather = await getWeather();

  const prompt = `Cheerful illustration for ${today} morning, ${weather}, motivational, warm colors`;

  return await generateImage(prompt);
}

Project Visualization

You: "Visualize our new mobile app concept" Bot: [Generates mockup-style image based on project details in memory]

Infographics

You: "Create an infographic showing my task completion this week" Bot: [Generates visual summary of productivity stats]

Custom Avatars

You: "Generate an avatar for our team Slack channel, we're a fintech startup" Bot: [Generates professional fintech-themed avatar]

Cost Management

Track Usage

IMAGE_MONTHLY_BUDGET=5.00  # GBP
IMAGE_DAILY_LIMIT=10
NOTIFY_AT_BUDGET_PERCENT=80

Bot at 80%: "Monthly image budget is 80% used (£4.00/£5.00). 20 images remaining."

Cost Per Image

| Provider | Size | Quality | Cost | |----------|------|---------|------| | DALL-E 3 | 1024x1024 | Standard | $0.040 | | DALL-E 3 | 1024x1024 | HD | $0.080 | | DALL-E 3 | 1792x1024 | HD | $0.120 |

Typical Monthly Costs

| Usage | Images/month | Cost | |-------|--------------|------| | Light | 50 | ~£2 | | Moderate | 150 | ~£6 | | Heavy | 500 | ~£20 |

Troubleshooting

"Content policy violation"

Some prompts are blocked. Try:

More generic descriptions
Remove specific names/brands
Avoid sensitive topics

Slow generation

DALL-E 3 HD takes longer
Use standard quality for previews
Check API status

Poor quality results

Use more detailed prompts
Enable prompt enhancement
Try different styles

Image not sending

# Check bot logs
pm2 logs openclaw | grep -i "image\|dall"

Verify file size limits for Telegram/Discord
Check API key validity

Security Notes

Don't Generate

Copyrighted characters (without permission)
Real people's faces
Harmful or illegal content
Content that violates platform terms

Logging

# Log image prompts for review
LOG_IMAGE_PROMPTS=true
IMAGE_LOG_PATH=/var/log/image-generation.log

Related Guides

Need Help?

Image generation setup involves API configuration and cost management. Our setup service includes image generation integration with your preferred provider.