Premium members can use custom prompts to control and personalise their media generations.
A custom prompt can add detail to your selected presets, refine them or replace optional choices such as the outfit and scene. The best prompts describe one clear, finished image using details that can be seen.
Custom prompts can contain up to 500 characters.
First, select your creator and the pose that most closely matches the body position or action you want.
The pose provides the image’s core structure, including whether the creator is standing, sitting, kneeling, squatting or lying down.
You can then select an optional:
Scene
Outfit
Modifier
Second modifier
These optional presets provide useful starting details. Your custom prompt can change them or fill in details they do not cover.
Custom prompts are useful for controlling:
Outfit style, colour, material, pattern and fit
Scene, location and background
Lighting, weather and time of day
Expression and gaze
Hair styling
Arms and hands
Props and nearby objects
Camera angle and framing
Mood and atmosphere
Small changes to body positioning
For example, if you select the Underwear outfit, you could refine it with:
Red floral lace lingerie with scalloped trim and delicate matching stockings.The custom prompt provides the more specific outfit details.
You can also replace an optional preset completely. If you select a beach scene but describe a mechanic’s workshop in your custom prompt, the workshop becomes the intended location.
If you already know exactly what outfit or scene you want, you can leave that preset empty and describe it directly.
A custom prompt cannot replace the selected pose’s core structure.
Protected pose details include:
The main standing, sitting, kneeling or lying position
Torso and lower-body direction
Weight-bearing arms or legs
Required support surfaces
Important physical contact points
The pose’s core action
For example, selecting a kneeling pose and asking the creator to stand against a wall creates a conflict. The creator will remain kneeling, although compatible details—such as the wall, outfit, lighting, expression or hand placement—may still be used.
For a standing result, select a standing pose first.
Smaller pose details are more flexible. You can usually change:
Free arm and hand placement
Head angle
Gaze and expression
Hair position
A slight lean or arch
Camera treatment
These changes will be followed when they remain physically compatible with the selected pose. A hand needed to support the creator’s body cannot also hold a prop, so the prop or hand request may be adjusted.
FanFaves combines your choices in this order:
Creator permissions, generation mode and platform rules
Creator identity and face visibility
The selected pose’s core structure
Your custom prompt
Outfit, scene and modifier presets for details you did not specify
This means your custom prompt has priority over conflicting outfit, scene and modifier descriptions, but it cannot override the selected pose’s foundation or the active content mode.
For the most predictable result, choose a pose that already supports your idea and avoid unnecessary contradictions.
Write what should be visible instead of giving the generator a list of commands.
You do not need to use the creator’s name or repeat their physical appearance. FanFaves already manages the creator’s identity.
A useful prompt structure is:
Outfit + location + expression or action + camera or lighting + special detail
For example:
Black floral lace lingerie in a modern hotel bedroom, sitting at the edge of the bed, looking into the camera with a confident smile and warm bedside lighting.This is clearer than:
Change her outfit, put her somewhere sexy, move her around, change the camera and make the lighting look better.Clear visual details work better than long stories or lists of possibilities.
Useful details include:
“Black floral lace” instead of “sexy underwear”
“Warm bedside lighting” instead of “good lighting”
“Rainy city street at night” instead of “outside”
“One hand resting in her hair” instead of “a different pose”
“Low camera angle” instead of “make it dramatic”
Avoid using alternatives such as:
Either
Maybe
Red or black
Sitting or standing
Bedroom or hotel
Surprise me
The generator must resolve your prompt into one image. Giving it several alternatives makes the result less predictable.
Describe what you want to see rather than concentrating on what should not appear.
Instead of:
No crowds, no bright sunlight and don’t make it look busy.
Try:
A quiet private beach at sunset with soft golden light and an uncluttered background.
Positive descriptions give the generator a clearer finished result to create.
Modifiers are designed for specific visual changes and may be more reliable than describing the same effect only in text.
When a suitable modifier exists:
Select the modifier.
Use your custom prompt to describe its colour, styling, intensity or surrounding details.
Make sure it works with your pose, outfit and scene.
You can select up to two modifiers. Adding more details does not always improve the image, so use only the modifiers that are important to your idea.
Consider how your requested image will fit inside the selected frame.
4:5 works well for portraits and social posts.
9:16 is useful for standing and full-body compositions.
1:1 suits closer, balanced portraits.
5:4 provides more room for the environment.
16:9 works well for wide or cinematic scenes.
A wide environment in a narrow portrait ratio may be cropped or simplified. A detailed full-body pose usually benefits from a taller frame.
Custom prompts cannot override:
The creator’s enabled content tiers
Suggestive or Explicit mode limits
Creator identity requirements
FanFaves platform rules
Legal and consent requirements
A Suggestive generation cannot be turned into an Explicit generation through a custom prompt. Switch to Explicit mode first, where that mode is supported.
All depicted people must be consenting adults aged 18 or older. Requests involving minors, non-consensual activity, incest, incapacitated people, animals or other prohibited content will not be generated.
White floral lace lingerie with scalloped edges, delicate embroidery and a matching satin robe resting loosely over her shoulders.
Inside a dimly lit mechanic’s workshop, leaning beside a classic car with warm overhead lights, metal tools and subtle grease marks on her skin.
Looking directly into the camera with a playful smile, one hand resting in her hair and the other relaxed beside her hip.
Red velvet mini dress in an elegant hotel bar, seated with one hand holding a cocktail, confident eye contact, warm golden lighting and softly blurred city lights behind her.
Your prompt probably conflicted with the selected pose. Choose a pose with the correct standing, sitting, kneeling or lying structure.
Describe one specific outfit or location and remove competing alternatives. You can also clear the optional outfit or scene preset when replacing it completely.
Make sure a free hand is available to hold it. If both hands support the pose or perform its core action, the generator may remove the prop to preserve realistic anatomy.
Simplify the request. Remove unnecessary modifiers and describe one outfit, one location, one action and one mood.
Before generating, ask:
Does my selected pose match the body position I want?
Have I described one finished image?
Did I choose one outfit and one location?
Are my modifiers compatible?
Are the requested hand and prop placements physically possible?
Does my aspect ratio suit the composition?
Is my prompt within the selected content mode?
A focused prompt with compatible presets will usually produce the strongest result.