Image Generation
With the Image Generation tool, the AI assistant can create original images based on text descriptions. This is particularly useful for communications departments and policy documents.
Capabilities
- Visualizations: Create illustrations for policy documents
- Infographics: Generate visual representations of data
- Communications material: Create images for public information
- Concept visualizations: Show what projects could look like
This article is about the tool the AI can use from a regular chat. If you want to open image mode directly and choose settings such as model, size, style, color preference, variants, and reference images yourself, read Text to Image.
For images created with this tool in a regular chat, you can use the model button on the answer to regenerate with a different image model. In direct image mode, you manage the model and other options beforehand through the input field.
Example Usage
"Create an illustration symbolizing collaboration between municipality and residents on neighborhood sustainability"
"Generate an infographic-style image about waste separation"
"Create a visualization of a green, accessible playground"
Supported Models
The following AI models can generate images:
| Model | Quality | Speed |
|---|---|---|
| GPT Image / Nano Banana / FLUX | Depends on selection | Depends on selection |
| GPT-4o | High | Fast |
| GPT-Image | High | Fast |
Considerations
AI-generated images are suitable for internal documents and concepts. For external publication, consider:
- Checking images for unintended elements
- Ensuring the style matches your visual identity
- Indicating that the image is AI-generated (if relevant)
Enable
- Go to Admin → Settings → Tools
- Find Image Generation
- Enable the tool
Combine this tool with clear descriptions of your brand colors for more consistent results.
