> ## Documentation Index
> Fetch the complete documentation index at: https://learn.getodin.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Image Generation Toolkit

> Generate high-fidelity images on demand using DALL-E 3

This toolkit gives the agent the ability to act as a creative director, generating high-fidelity images on demand using DALL-E 3. It transforms text descriptions into visual assets instantly. The agent can create original artwork, illustrations, diagrams, and photorealistic images based on natural language descriptions.

## Configuration Options

<img src="https://mintcdn.com/odinai/JI14YIDoIJKTxHdQ/img/agents-toolkits/13.jpg?fit=max&auto=format&n=JI14YIDoIJKTxHdQ&q=85&s=f2afb9e6cde6f5347f49aa41fee7af4e" alt="Image Generation Configuration panel with DALL-E settings" width="1280" height="630" data-path="img/agents-toolkits/13.jpg" />

The Image Generation Toolkit configuration panel allows you to set default preferences for image generation:

* **Provider**: DALL-E (OpenAI) - The AI image generation provider.
* **Model**: DALL-E 3 - The model version used for image generation.
* **Default Image Size**: Choose from 1024x1024 (Square), 1024x1792 (Portrait), or 1792x1024 (Landscape).
* **Default Quality**: Standard (faster, lower cost) or HD (more detailed, higher cost).
* **Maximum Images Per Request**: Generate 1-4 images at once (set range 1-4).

## Image Size Options

| Size                  | Description            | Use Case                   |
| --------------------- | ---------------------- | -------------------------- |
| 1024x1024 (Square)    | Standard square format | Avatars, icons             |
| 1024x1792 (Portrait)  | Vertical orientation   | Phone wallpapers, posters  |
| 1792x1024 (Landscape) | Horizontal orientation | Desktop wallpapers, slides |

## Available Tools

The Image Generation Toolkit provides one tool that can be enabled or disabled:

* **Generate Images**: Generate one or more images from a natural language prompt using DALL-E. This tool can be configured to require user confirmation before execution.

## Storage & Display

* **Automatic Upload**: Generated images are automatically uploaded to your project storage.
* **Persistent URLs**: Images receive permanent URLs for reliable access.
* **Chat Integration**: Images are attached to chat messages and displayed immediately in responses.

## Prompting Best Practices

* Be specific about style, lighting, and composition (e.g., "shot from above with natural sunlight").
* Use descriptive adjectives (e.g., "cinematic lighting," "hyper-realistic," "minimalist," "vintage").
* Avoid vague terms like "cool" or "nice" - instead describe exactly what you want.
* Iterate on prompts based on generated results - refine and adjust for better output.
* Specify perspective and framing (e.g., "close-up portrait," "wide-angle landscape," "bird's eye view").
* Include art style references when helpful (e.g., "in the style of Art Nouveau," "watercolor painting").
* Mention mood and atmosphere (e.g., "mysterious," "cheerful," "dramatic," "serene").
* For technical diagrams, be precise about layout and elements required.

## Use Cases

* **Marketing**: "Generate 5 concepts for our new sneaker ad campaign."
* **Storyboarding**: "Create a scene showing a cyberpunk city in rain."
* **Web Design**: "Generate a placeholder hero image for a gardening website."
* **Presentations**: "Create an illustration representing 'Cloud Security' for my slide deck."