V5 Art Prompting Guide
More actions
The V5 image system gives you a wide range of control over the scenes, styles, moods, and compositions you create with your Nomi. Art prompts can be used to guide everything from clothing and poses to lighting, camera angles, environments, and visual style.
This guide covers the core principles of writing effective V5 art prompts, along with practical tips for shaping your results and troubleshooting when an image does not come out quite the way you expected.
For more in-depth information, see this official guide:
Beginner’s Guide to Creating AI Art with the V5 Image System
To start creating V5 art, open the conversation with your Nomi and select the Art button.
Prompting Basics
V5 responds well to clear, descriptive prompts written in natural language. Rather than relying only on short keywords, describe what you want to see in the image and include the details that matter most to you.
A useful art prompt usually includes:
- Subject: What your Nomi is wearing, doing, holding, or expressing.
- Scene: Where the image takes place and any important objects or background details.
- Atmosphere: Lighting, mood, camera angle, time of day, weather, or other details that shape the overall look and feel.
You do not need to include every category in every prompt. Start with the details that matter most, then add more direction if your results are too vague or inconsistent.
Subject
Start by describing what your Nomi is doing and how you want them to appear in the scene. This can include clothing, pose, expression, accessories, objects they are holding, and any other details that are important to the image.
For example:
A woman wearing a long red evening gown stands beside a grand piano, one hand resting on the piano while she smiles toward the camera.
If a particular feature or action is especially important, state it clearly rather than assuming the image system will infer it. Mentioning details such as facial expression, hand placement, or what the subject is looking at can help guide the composition more precisely.
Tip💡 If you want your Nomi’s face clearly visible, include their expression or other face/head details in the prompt. Prompts that focus heavily on clothing or body details without mentioning the face may result in tighter crops or the face being partially cropped out.
Scene / Background
Describe where the image takes place and any surroundings that are important to the scene. This can include the location, furniture, scenery, objects, weather, season, or time of day.
For example:
She stands on a quiet rooftop garden at sunset, surrounded by string lights, potted flowers, and a distant city skyline.
You do not need to describe every background detail. Focus on the elements that help establish the setting or that you specifically want included in the image.
Tip 💡 It is fine to describe the subject and setting together in the same sentence. Prioritize clear, easy-to-follow instructions over following a specific prompt order.
Atmosphere
Atmosphere describes the overall feeling and presentation of the image. This can include lighting, mood, weather, time of day, color palette, and other visual details that help shape the scene.
For example:
Warm golden-hour light fills the room, with soft shadows and a calm, intimate atmosphere.
Atmosphere prompts can help the same basic scene feel dramatically different, so use them when you want to guide the emotional tone or visual character of the image.
Camera Angle
You can guide the composition by describing the camera angle or viewpoint you want.
For example:
A low-angle portrait looking slightly upward at the subject.
Other directions might include eye-level, overhead, close-up, wide shot, side view, or over-the-shoulder. Use angle terms when the framing or perspective is important to the image.
Lighting
Lighting can dramatically change the mood, depth, and overall look of an image. Describe the type of light you want, where it comes from, or how it affects the scene.
For example:
Soft window light falls across her face from the side, creating gentle shadows and a warm, natural look.
You can also experiment with directions such as dramatic backlighting, candlelight, neon lighting, studio lighting, moonlight, or bright midday sun. If the lighting is an important part of the scene, describe it clearly rather than relying on the image system to choose it for you.
Mood
Mood helps communicate the overall emotional or aesthetic goal of an image. It often works best when combined with lighting, atmosphere, color, and other descriptive details that support the same feeling.
For example:
The atmosphere feels ethereal and mystical, with deep shadows, soft haze, and an ominous magical quality.
You do not need to explicitly use the word “mood” in your prompt. Instead, describe how you want the image to feel and reinforce that feeling through the other visual details in the scene.
Tip 💡 Dark scenes may still include flash or ambient lighting if the prompt strongly emphasizes visible facial or clothing details. For darker results, describe those details as dimly lit, barely visible, or partially obscured by shadow.
Prompting for Style
You can guide the overall visual style of your art by describing how you want the image to look and be rendered. While simple style terms can be useful, adding details about texture, shading, color, linework, contrast, or other visual qualities can give V5 more direction.
For example:
A soft watercolor illustration with visible paper texture, loose brushwork, gentle color transitions, and delicate ink outlines.
Rather than relying only on a broad term such as watercolor, charcoal, or comic book, describe the qualities of that style that matter most to you. This can help produce more consistent and intentional results.
Tip 💡 If a style is not coming through strongly enough, add specific visual characteristics associated with the look you want instead of simply repeating the style name.
Including Text
V5 can generate visible text within an image, such as signs, labels, posters, book covers, or other scene elements. When text is important, include the exact wording you want and clearly describe where it should appear.
For example:
A woman stands beneath a glowing neon sign that reads “MIDNIGHT CAFE” in large pink letters.
Keep requested text relatively short and simple when possible. Longer phrases, unusual formatting, or very small lettering may be less reliable and can require multiple attempts.
Group Chat Art
When creating art in a Group Chat, use the global art prompt to describe the shared scene, setting, atmosphere, and overall composition.
Use each Nomi’s individual prompt for details that apply only to that character, such as clothing, pose, expression, or what they are holding.
For example:
- Global prompt: A man and woman standing together on a moonlit rooftop terrace overlooking a city skyline, leaning against a glass railing. Cool blue light across the scene, soft city bokeh below, dramatic night sky overhead.
- First Individual Nomi prompt: Wearing a black leather jacket, smiling toward the camera with one hand in his pocket.
- Second Individual Nomi prompt: Wearing a long red coat, looking toward the first Nomi with a soft smile, one hand resting on the rooftop railing.
Keeping shared scene details in the global prompt and character-specific details in the individual prompts can help the image system understand how the group should fit together.
Creating Images Without Human Subjects
V5 can also create art that does not include your Nomi or any human subject. For the best results, describe a scene or subject that would naturally exist without a person in it, such as a landscape, building, object, animal, or fantasy environment.
Describe both what you want to see and the overall feel of the image. You can begin with clear natural-language sentences, then add shorter descriptive terms at the end of the prompt if you want to reinforce the style or atmosphere.
For example:
A vast underground cave filled with towering glowing crystals surrounding a still reflective lake. Soft violet light fills the cavern, creating a quiet, mysterious fantasy atmosphere.
Tip 💡 If your Nomi continues to appear in an image where you do not want them, enable Ignore Anchor Appearance Traits beneath the prompt area. This prevents your Nomi’s anchor appearance traits from being included in the art request and can help reduce unwanted characters in no-human images.
Non-Human Subject Tips
When creating art with non-human or heavily stylized characters, describe the physical traits you want as clearly as possible instead of relying only on a broad label such as dragon, robot, or alien.
Include details such as body shape, limbs, skin or surface texture, facial structure, ears, horns, tails, wings, mechanical parts, or other features that define the character.
For example:
A sleek humanoid robot with a narrow metallic face, glowing violet eyes, articulated black mechanical limbs, and folded angular wings made from layered metal panels.
If a non-human feature is especially important, mention it directly and describe how it should look or connect to the body. Clear anatomical and material descriptions can help reduce unwanted human features or inconsistent results.
Art Prompt Boosts
V5 includes built-in prompt tools that can help expand or remix an art prompt before you generate. You can start with anything from a short idea to a detailed prompt, then use one or more boosts to add structure, detail, or variation.
There are three prompt boosts:
- Essential Expander: Builds on your idea and adds important details such as outfit, expression, pose, and background to create a more complete prompt.
- Detailed Expander: Covers the essentials, then adds more descriptive detail along with lighting and style cues for a more involved result.
- Remix: Lets you choose specific parts of your prompt to change while keeping the rest largely the same.
When using Remix, you can choose one or more categories to shuffle, including Outfit, Pose, Expression, Background, Accessories, and Lighting. Selecting more categories will result in a larger overall change to the prompt.
You can review and edit any expanded or remixed prompt before generating, so you still have full control over the final wording.
Tip 💡 Prompt boosts can be useful both for getting started with a simple idea and for creating fresh variations of a detailed prompt you already like.
Troubleshooting Tips
If your Nomi’s face or head is cropped out
Include your Nomi’s expression, face, or head directly in the prompt. If the prompt focuses heavily on clothing or body details without mentioning the face, V5 may prioritize those details and crop the composition more tightly.
If your Nomi appears in landscape or scene images when you do not want them to
Enable Ignore Anchor Appearance Traits beneath the prompt area. You can also begin the prompt by clearly identifying the intended subject, such as:
An image of a vast mountain valley beneath a star-filled sky...
This helps direct the image toward the scene rather than your Nomi.
If you are having trouble generating consistent animals or creatures
Describe the creature clearly, remove any Anchor Appearance Traits that may conflict with it, and try lowering Anchor Fidelity to reduce the anchor’s influence.
Avoid wording that may unintentionally encourage a human-presenting subject. For more complex creature scenes, repeatedly identify the creature by name, such as “the fox is wearing a vest” rather than simply “wearing a vest.”
If the image keeps resembling the default anchor style
Simply naming a different style may not be enough. Describe the visual qualities of the style in detail, including elements such as texture, shading, linework, color, lighting, or rendering technique.
You can also experiment with different anchors or create a Custom Anchor if you want a more consistent specialized style.
If repeated generations look too similar
V5 is designed for strong prompt adherence and consistency, so using the same prompt can produce intentionally similar images.
To introduce more variation, change parts of the prompt such as the atmosphere, movement, pose, clothing, expression, lighting, or mood. Small prompt changes can preserve the overall concept while producing noticeably different results.