This article will show you how to easily convert text to images using the “Text to Image” node in AI Content Labs. You’ll see how this node fits into workflows and provides options for customizing your creations, from choosing the AI model to how the results are displayed.
What is the Text to Image Node and What is it Used For?
The Text to Image node is responsible for transforming any text definition (prompt) into one or more images generated by Artificial Intelligence. For example, you could take an Input Node describing “a sunset on the beach” and pass it to the node to produce images of that landscape. A key use is in workflows where you need to illustrate textual content automatically and quickly.
The image generation source varies: you can choose between different models such as DALL·E, Stable Diffusion or Flux, among others. For a complete guide, check the list of available text-to-image models.
Settings
Although the exact appearance depends on the model you choose, the Text to Image node usually displays the following main settings:
Model (Source)
Here you select the AI provider. For example, in the image you can see “Leonardo Phoenix” or “dall-e-3” as possible sources. Some models offer additional configurations, while others simplify the interface.
Prompt and Negative Prompt
- Prompt: The main phrase or idea that will be converted into an image. It can be a direct input (e.g., “a forest in autumn”) or linked to the output of another node.
- Negative Prompt (optional): Useful in some providers to eliminate unwanted elements in the generation. Not all models accept it.
Additional Model Options
Each provider may display different settings. Some examples:
- Aspect Ratio: Controls the aspect ratio of the image (e.g., 21:9).
- Dimensions (Width/Height): Adjusts the image size in pixels.
- Number of Images: Indicates how many images to generate per request.
- Styles and Filters: Some models like Leonardo Phoenix include style options (3D Render, photographic, etc.) or filters like “Alchemy”, “Ultra” or “Enhance Prompt”.
- Seed: To reproduce the result with the same prompt and configuration.
- Safety Checker: Option to filter inappropriate content.
As you can see, each model has unique settings:
| Parameter | Example | Description |
|---|---|---|
| Contrast | Medium (Leonardo) | Increases or decreases color intensity. |
| Output Format | PNG or JPG | Defines the format of the resulting image. |
| Style UUID | “3D Render” | Predefined styles of certain providers. |
Output Options
Like all AI Content Labs nodes, it has common options to hide the output in the final result, exclude it from webhooks, or send it in HTML. Additionally, for the Text-to-Image node, we have:
- Markdown Format: Converts the image output into Markdown formatted links.
- Separator Pattern: If you generate more than one image, you can define a separator (e.g.,
###) to distinguish them.
Usage Tips
- Combine with Prompt Node: You can first refine or generate the text with a Prompt Node and then pass it directly to the Text to Image node.
- Use Negative Prompt wisely: In models that allow it, specify elements you want to exclude (e.g., “no characters” or “no fade effect”) to achieve more precise results.
- Take advantage of the Separator Pattern: If you generate multiple images, specify it to then use a Text Splitter and work with each one independently (e.g., increase its resolution with the Image Tools node).
- Experiment with styles: Each model has its unique touch. Try different providers or parameters like “Ultra”, “Alchemy” or “3D Render” to vary the result.
- Control the size: Adjust the dimensions and aspect ratio to what you need (a blog template, a presentation, a banner, etc.).
With the Text to Image node, you can easily transform your ideas into illustrations, adjusting parameters according to the AI model you choose. Experimenting with different settings allows you to find the perfect balance between creativity, consistency, and quality. Enjoy the versatility offered by this node in your workflows and get the most out of visual artificial intelligence.
