Use each control for a distinct choice
The prompt describes the desired picture. A negative prompt lists unwanted details. Style selects a visual treatment, image count concerns quantity and creativity changes another aspect of the setup. Read each field rather than putting every instruction into one long sentence.
Keep an eye on any allowance or cost shown for the selected action. The account's text-message access does not establish what an image request includes.
| Control | Useful input |
|---|---|
| Prompt | Adult subject, clothing, location, framing and light |
| Negative prompt | Details that should be excluded |
| Style | The intended realistic or illustrated treatment |
| Quantity | How many results you want to compare |
| Creativity | A setting to vary deliberately between attempts |
Keep the scene visually consistent
Choose one adult fictional subject and describe the view: portrait, full figure or a wider scene. Add clothing, a location and one lighting direction. Conflicting camera positions or several incompatible art styles can make the intended result unclear.
Use invented identities and material you have permission to use. Do not supply private photographs or another person's intimate likeness without consent.
An adult fictional traveler in a raincoat beneath a station awning, waist-up portrait, early morning light, wet platform in the background, realistic photography style.
Change one meaningful detail at a time
Keep a useful prompt as the comparison point. Change framing, lighting or one style setting, then compare what that change did. If you change every choice at once, it is harder to identify which instruction improved the image.
When a still picture is ready for motion, use the relevant image-to-video route if available. A motion description serves a different purpose from the original still-image prompt.
Questions and answers
How should I write a GirlfriendGPT image prompt?
Name an adult fictional subject, describe clothing and setting, choose a framing and specify the light. Keep the visual style consistent. Use the negative prompt for unwanted elements instead of mixing contradictory requests into the main description.
Does a picture use the same allowance as a chat reply?
Not necessarily. Read the allowance and cost for the image action in the selected tool. Text access, included credits and a media request can have different conditions.
How do I give a still image motion?
Use the image-to-video option where it is offered. Start from the chosen picture and describe a simple movement or camera change. The result is a generated clip rather than a live camera session.
Are image tools included in free access?
Use the conditions shown for your account and the selected request. Starting a text conversation without a subscription does not establish that every image action is included at no cost.
What belongs in a negative image prompt?
List visible elements you want excluded, such as text, a watermark, a cropped face or an unwanted object. Keep the desired subject and scene in the main prompt. Avoid asking for the same detail in one field and excluding it in the other.