Choose a character and image settings
- Open Generate Image and select Pick a Character.
- Use the available Outfit, Action, Pose, Accessories and Scene suggestions, or describe the result you want.
- Choose the number of images. Review the current quantity options and token charge before confirming.
- Select Generate Image.
- Find completed results in Generated Images. Save the result you want rather than paying for another generation because a download was blocked.
A practical prompt structure
Try this order: subject, pose, clothing, setting, lighting, framing. For example:
An adult character in a dark denim jacket, seated at a cafe window with a cup on the table. Soft daylight, relaxed expression, waist-up portrait.
For a wider scene, replace the framing with full-length view with the cafe entrance visible
. Changing one element makes it easier to see which instruction affected the result.
Refine a result
If the picture misses a detail, shorten the prompt and emphasize the most important instruction. Several conflicting poses, outfits or locations can make a request harder to follow. Appearance and clothing details may vary between attempts; a repeated prompt is not a guarantee of an identical image.
Generation is different from editing an existing file. Only use an edit, upload or image-to-video tool when that control is actually offered in your account.
Editing a prompt or an existing file
Changing a prompt and generating again creates a new result; it does not edit the file you already saved. Generate Image is a character-and-prompt workflow, not an upload instruction. Use a separate editor only when that tool is offered for the selected content, and check its own controls and cost.
Control image responses in chat
Premium users can open the Ask dropdown in a character conversation and switch Image Generation off or on. When it is off, image-request options such as Show me
are removed and image-triggered responses are disabled. The setting is remembered across sessions on supported desktop and mobile interfaces. This toggle does not make other token-based actions free.
Check the action before confirming
Review the token charge, number of images and selected settings. Save a result using the available download control and your browser's permissions. If a download fails, check the browser's file permissions before creating another paid image.
For a failed video save, follow the media download checks. Image-to-video is a separate action with its own settings and token cost.