Plan an image before generating it
A focused prompt gives an image request a clear subject. Describe the adult character and one scene, then add the visual details that matter most.
Give the request a simple structure
- Identify the fictional adult character and the appearance you want to retain.
- Specify one outfit, action and location.
- Add the composition and lighting, such as a waist-up portrait in soft window light.
- Review the selected options and displayed token cost before submitting.
An example to build on
A 30-year-old botanist in a green linen jacket, standing beside a greenhouse workbench, holding a small plant pot, waist-up composition, soft morning window light.
If you want a different result, change one important detail at a time. Replacing the setting, outfit and lighting together makes it harder to tell which change affected the image.
Check the result and your balance
Wait for the submitted request to finish before starting the same request again. Open the completed result and use the download option available for that file.
An image request can use tokens independently of your text allowance. If a charged request fails, note its approximate time and visible error for support. Do not assume a repeated click is free.
Questions and answers
Can I turn every Candy AI image into a video?
No. Image-to-video depends on an eligible result and an available conversion control. A saved image alone does not guarantee that option.
Will adding more words always improve an image prompt?
No. Compatible details are more useful than a long list of conflicting poses, outfits or lighting conditions. Start with one subject and scene, then refine the result.