Article Generation
Image Generation
Creating visuals for your articles
Why Custom Images?
Original images set your content apart. Instead of using the same stock photos everyone else uses, Dipflow creates unique visuals tailored to your specific article.
Benefits of AI-generated images:
- Unique visuals that match your exact content
- No licensing concerns or attribution requirements
- Consistent style across all your articles
- No need to search through stock photo sites
Types of Images
Dipflow generates two types of images:
Main image (featured image)
The primary visual for your article. This appears at the top of the content, in social shares, and in article previews. It's designed to capture attention and represent the article's topic.
Inline images
Additional visuals placed throughout the article content. These break up long text sections and illustrate specific points. The number of inline images depends on your article length and structure.
The Generation Process
Image generation happens in two steps:
1. Visual planning
The AI analyzes your article and creates image prompts - detailed descriptions of what each image should contain. These prompts are shown to you for review.
2. Image creation
Using the prompts, the AI generates the actual images. Each image takes a few seconds to create. You can track progress as images appear in the preview.
Image generation uses credits. Each image (main or inline) costs credits to generate. Check the credits documentation for specific costs.
Managing Images
After images are generated, you can:
Review and accept
Look at each generated image. If it works for your content, keep it and move forward.
Regenerate
If an image doesn't work, regenerate it. The AI will create a new variation based on the same prompt. Regeneration uses additional credits.
Edit prompts
Modify the image prompt before regenerating for more control over the output. Be specific about what you want to see.
Skip images
If you prefer to use your own images or don't want visuals for a particular section, you can skip image generation for those placements.
For best results, review image prompts before generation. Small tweaks to prompts can significantly improve the output.