AI Prompts: How to Write Prompts for Images and Videos + 22 Ready-to-Use Examples

AI prompts determine what appears in an AI-generated image or video: the subject, lighting, camera angle, composition, and action. If you type something vague like "a beautiful woman in a forest," the model has to make most of the creative decisions itself. It chooses the person's appearance, clothing, surroundings, lighting, and framing. The result might look impressive, but it probably won't match what you had in mind.

A good prompt doesn't have to be long. What matters is replacing vague descriptions with specific visual instructions. Instead of "a beautiful woman in a forest," try "a woman wearing a beige coat, standing on a forest path after the rain, soft morning light, eye-level camera angle, slightly blurred background." For video generation, there's another dimension to consider: time. You need to describe what moves, how it moves, and what the camera does.

Below, you'll find a practical formula for writing AI prompts, seven essential rules, and 22 ready-to-use prompts for images and videos. You can copy any example, customize a few details, and try it directly in Neuromia.

What Is an AI Prompt? A Simple Explanation

An AI prompt is an instruction you give an artificial intelligence model. For image generation, it describes the picture you want to create. For video generation, it describes both the scene and the actions that happen over time.

A prompt can be a single sentence or a detailed description covering lighting, composition, camera angles, and movement. The more clearly your instructions communicate your idea, the fewer creative decisions the model has to make on its own.

A prompt is not the same thing as a generator's technical settings. The text describes what should appear in the image or video, while settings such as aspect ratio, resolution, and video duration control the output format when the selected model supports them.

You also don't need to learn complicated commands or special syntax. Most image and video generators accept ordinary natural language. The challenge is knowing which details to include and which ones to leave out.

The AI Image Prompt Formula: What to Include

A reliable structure for image generation prompts is:

Subject → Action or pose → Environment → Style → Lighting → Camera angle and composition → Details → Format.

This isn't a strict command format. You don't need arrows, labels, or special separators in your actual prompt. The formula is simply a checklist that helps you describe a scene without forgetting essential details.

The Eight Elements of a Good Image Prompt

Element	What to describe	Example
Subject	Who or what appears in the image	A woman with short dark hair
Action or pose	What the subject is doing	Sitting sideways on a wooden chair
Environment	Where the scene takes place	A bright studio with a beige backdrop
Style and technique	How the result should look	Realistic portrait photography
Lighting	The direction and quality of light	Large soft light source from the left
Camera and composition	The viewpoint and subject placement	Waist-up portrait, eye-level camera
Details	Important textures and visual features	Natural skin texture, individual hair strands
Format	Image shape and intended use	Vertical 4:5 composition

You don't need all eight elements in every prompt.

A simple product photograph might only need a subject, background, lighting setup, and camera angle. A more complex scene involving several characters or objects will benefit from additional detail.

Complete image prompt example:

Realistic studio portrait of a young adult woman with short chestnut-brown hair. She is sitting sideways on a wooden chair, looking directly at the camera. Clean light-beige studio background with no distracting objects. A large soft light source from the left creates gentle shadows across her face. Waist-up framing, eye-level camera, natural portrait perspective, realistic skin texture, calm facial expression. Vertical 4:5 composition.

Every sentence has a purpose. The model receives clear information about the subject, environment, lighting, and photographic setup.

You can test this formula using AI image generation in Neuromia. The image editor includes a "Prompt Examples" section where clicking an example automatically inserts it into the prompt field.

From a Weak Prompt to a Strong One: Three Steps

Imagine you want a realistic photograph of a woman in an autumn forest.

Step 1. The weak prompt

A beautiful woman in a forest.

The problem is that almost everything is undefined. What kind of forest? What is she wearing? Is it morning or evening? Should the camera show her face or her entire body?

Even the word "beautiful" leaves considerable room for interpretation.

Step 2. Add the subject, environment, and action

A young adult woman with long chestnut-brown hair wearing a light-colored coat stands on a path in an autumn forest. Yellow and orange leaves surround her, with tall trees in the background. She is looking slightly away from the camera.

Now the scene has a recognizable subject and setting. However, the model still has to decide how to frame and light the shot.

Step 3. Add photographic direction

Photorealistic autumn portrait of a young adult woman with long chestnut-brown hair. She is wearing a light-beige coat and a knitted scarf, standing on a narrow forest path surrounded by golden birch trees. She looks slightly to the right of the camera. Early morning, soft sunlight filtering through the leaves, subtle mist in the background. Waist-up framing, eye-level camera, naturally blurred background, warm color palette, realistic skin and hair textures. Vertical 4:5 photograph.

The third version doesn't guarantee an exact result, but it gives the model a much clearer target.

The main lesson is simple: don't just add more flattering adjectives. Add details that describe what the camera should actually see.

How to Write Better AI Prompts: 7 Practical Rules

1. Replace Vague Adjectives with Specific Details

Words like "beautiful," "amazing," "luxurious," and "atmospheric" are open to interpretation.

Instead of:

"A luxurious bedroom."

Try:

"A spacious bedroom with a floor-to-ceiling window, a bed with a tall upholstered headboard, light stone surfaces, and warm indirect lighting."

Instead of:

"A delicious-looking dessert."

Try:

"A slice of chocolate cake with glossy chocolate ganache, fresh raspberries, and clearly visible layers of cream."

Specific details give the model something concrete to reproduce.

2. Put the Main Subject First

Start by explaining what you want to create.

Less effective:

Warm lighting, soft shadows, high detail, beautiful bokeh, a woman.

Better:

Portrait of a woman wearing a red sweater beside a window. Soft side lighting, blurred background, close-up composition.

Describe the subject and scene first, then add visual refinements.

This is particularly helpful when working with models that have relatively short prompt-length limits.

3. Keep One Image Focused on One Moment

Don't try to squeeze an entire story into a single photograph.

For example:

"A person wakes up, eats breakfast, goes to work, and returns home in the evening."

That describes multiple moments rather than one clearly defined image.

A better image prompt would be:

A person sitting at a wooden table beside a window, with a cup of coffee and an open book in front of them. Soft morning light.

If you need to show several events, create separate images or consider video generation.

4. Describe What Can Actually Be Seen

Abstract ideas are harder to visualize than physical actions and expressions.

For example, "a woman thinking about her future" doesn't tell the model how that thought should appear in the picture.

Instead, describe the visible behavior:

A woman sitting by a window, holding a closed notebook on her lap and looking into the distance. Calm, thoughtful expression, overcast daylight.

The same approach works for emotions. Rather than simply asking for a "happy person," describe a smile, relaxed shoulders, hand gestures, and the direction of their gaze.

5. Avoid Contradictory Instructions

Don't ask for a "dark nighttime scene with no lighting" and a "bright, sunny atmosphere" at the same time.

Another common conflict is requesting both an "extreme close-up of an eye" and "a full-body shot."

Choose what matters most.

If the face is the priority, use a close-up. If clothing and surroundings are important, choose a medium shot or full-body composition.

6. Describe Style Through Lighting and Camera Technique

The word "cinematic" can be useful, but it covers a huge range of visual styles.

More specific descriptions include:

• Soft, diffused natural light from a large window.
• Low-angle camera emphasizing the foreground.
• Shallow depth of field with a blurred background.
• Rim lighting outlining the subject's silhouette.
• Muted colors with moderate contrast.

For illustrations, use descriptions related to the medium instead: watercolor paper texture, flat vector shapes, pencil shading, or hand-drawn outlines.

You don't need to name an artist to describe a recognizable visual technique.

7. Change One Thing Between Attempts

Suppose your composition looks good, but the person's skin appears excessively smooth.

Don't rewrite the entire prompt.

Replace "smooth skin" with "natural skin texture, soft shadows, minimal retouching," then compare the next result.

If the composition is the problem, adjust only the camera angle or framing.

Changing one element at a time makes it easier to understand which instructions help.

Keep in mind that AI generation can produce different outputs even when the prompt remains unchanged. Evaluate the overall pattern rather than drawing conclusions from a single attempt.

13 Ready-to-Use AI Image Prompts

Here are 13 image prompts covering portraits, fashion photography, product shots, food, interiors, landscapes, illustrations, and more.

Each example is ready to copy and customize. You can use them as starting points rather than treating them as universally perfect prompts for every model.

Portraits and AI Photoshoot Prompts

1. Realistic Profile Picture or Avatar

Photorealistic portrait of a 30-year-old adult man with short dark hair wearing a plain graphite-gray T-shirt. Neutral, confident facial expression, looking directly into the camera. Light-gray studio background, soft lighting from both sides, natural skin texture, subtle catchlights in the eyes. Close-up head-and-shoulders framing, symmetrical composition, square 1:1 format.

What to change: Age, appearance, clothing, and background color. For a professional profile picture, replace the T-shirt with a shirt.

2. Studio Fashion Photoshoot

Realistic editorial fashion photograph of an adult model with short blonde hair wearing a flowing black dress. The model stands beside a light-gray wall, one hand resting on her waist, looking past the camera. Large soft light source from the right, carefully controlled shadows, minimalist studio setting. Full-body composition, camera slightly below eye level, understated color grading, vertical 3:4 format.

What to change: Outfit, background color, pose, and lighting. For a more dramatic look, request a narrow directional light source.

3. Outdoor Autumn Photoshoot

Photorealistic photograph of an adult woman wearing a brown trench coat and a light-colored scarf. She is walking along a narrow path in an old city park, gently holding her hair with one hand. Wet yellow leaves cover the ground, with blurred trees and benches in the background. Overcast afternoon after rainfall, soft natural light, knee-up framing, natural portrait perspective. Realistic colors, vertical 4:5 composition.

What to change: Season, clothing, location, and gaze direction.

Product Photography and Food

4. Clean E-commerce Product Photo

Professional product photograph of a matte white ceramic mug with no text or logos. The mug is centered against a seamless light-gray background, with its handle facing right. Soft studio lighting from the upper left, subtle contact shadow, clearly visible ceramic texture, rim, and shape. Three-quarter camera angle, minimalist composition, square 1:1 format.

What to change: Product type, material, color, and camera angle. For catalog photography, keep the background free of distracting objects.

5. Premium Perfume Advertising Image

Premium advertising photograph of a transparent glass perfume bottle without any branding, text, or logos. The bottle stands on a dark stone pedestal. Several suspended water droplets and curved translucent glass elements surround it. Deep graphite background, dramatic rim lighting, carefully controlled reflections, highly detailed glass and liquid textures. Centered composition, horizontal 3:2 format.

What to change: Liquid color, pedestal material, background, and decorative elements.

6. Appetizing Food Photography

Realistic restaurant food photograph of freshly prepared pasta with tomato sauce, basil leaves, and grated cheese served on a shallow ceramic plate. The plate rests on a dark wooden table beside a linen napkin and fork. Soft daylight coming through a window on the left, natural highlights on the sauce, visible ingredient textures. 45-degree camera angle, sharp focus on the food, horizontal 4:3 composition.

What to change: Dish, plate, background, and camera angle. For a restaurant menu, use consistent framing across different dishes.

Interiors, Landscapes, and Pets

7. Modern Living Room Interior

Photorealistic interior visualization of a cozy modern living room approximately 20 square meters in size. Light oak flooring, matte graphite cabinets, a muted plum-colored sofa, and a low TV console. Large window on the right, warm evening lighting, subtle integrated furniture lighting. Wide-angle interior photograph from the corner of the room, straight vertical lines, realistic materials, horizontal 16:9 format.

What to change: Room layout, finishes, furniture, and time of day. If you are redesigning an actual room, uploading a reference photograph can help communicate the existing layout.

8. Mountain Landscape at Sunrise

Photorealistic landscape of an alpine lake at sunrise. Calm, crystal-clear water reflecting sharp mountain peaks and a dense coniferous forest. A thin layer of mist floats above the lake, while the sky transitions through cool blue and soft pink tones. Wide establishing shot with a rocky shoreline in the foreground, natural lighting, highly detailed mountain textures and vegetation. Horizontal 16:9 composition.

What to change: Time of day, weather, landscape, and horizon placement.

9. Realistic Pet Portrait

Photorealistic portrait of a fluffy ginger cat with amber eyes. The cat is lying on a light-colored blanket beside a large window, looking directly into the camera. Soft morning light from the side, clearly visible individual strands of fur and whiskers, natural facial proportions. Close-up at the animal's eye level, shallow depth of field, calm beige background, square 1:1 format.

What to change: Breed, fur color, pose, and surroundings.

Fantasy Art, Illustrations, and Posters

10. Fantasy Character Illustration

Highly detailed digital fantasy illustration of an original forest guardian with silver hair and a dark-green cloak. She stands on an ancient stone bridge surrounded by enormous trees, holding a glowing crystal in one hand. Tiny golden particles float through the air, while morning mist rises beyond the bridge. Magical atmosphere, soft volumetric lighting, full-body composition, vertical 2:3 format.

What to change: Character appearance, handheld object, color palette, and environment.

11. Social Media Illustration

Modern flat vector illustration about working from home. A light-colored desk holds an unbranded laptop, a cup of coffee, a notebook, and a small plant. A window and simple geometric decorative shapes appear in the background. Clean outlines, soft beige, blue, and terracotta color palette, minimal fine details. Composition with generous empty space in the upper section for adding a headline later, vertical 4:5 format.

What to change: Topic, color palette, and placement of the empty space.

12. Vintage Film Photograph

Realistic photograph styled to resemble 1970s color film. A small roadside café with red chairs, a wooden counter, and large windows. A cup of coffee sits on a table, with a quiet street visible outside. Warm, slightly faded colors, subtle film grain, gentle contrast, natural exposure imperfections. Camera positioned at seated eye level, horizontal 3:2 composition.

What to change: Era, location, objects, and color treatment.

13. Minimalist Poster or Greeting Card

Minimalist graphic poster featuring a large orange circle representing the setting sun. Three overlapping dark-blue hill silhouettes occupy the lower section, against a cream-colored background with subtle printed-paper texture. Clean geometric composition, no gradients or unnecessary details. Short headline at the top reading "QUIET EVENING" in large, clear lettering. Vertical 2:3 format.

What to change: Headline, geometric shapes, and colors.

If accurate typography is essential, a more reliable workflow is to generate the background without text and add the lettering separately in an editor.

AI Prompts for Photo Editing and Enhancement

Generating a new image and editing an uploaded photograph are different tasks.

When editing an existing photo, the prompt should describe both the desired improvements and the elements that must remain unchanged.

For example, this is a useful AI photo enhancement prompt:

Improve the lighting and color balance of the uploaded photograph. Preserve the original composition, object proportions, and all important details. Reduce the color cast from artificial lighting, make the illumination look more natural, and gently recover details in the shadows. Do not add new objects or alter the scene's content.

This type of instruction is useful for product photography, interior images, and everyday pictures.

However, how accurately the original details are preserved depends on the selected model.

AI Video Prompts: What Makes Them Different?

Video prompts describe more than a scene's appearance. They also describe how the scene changes over time.

For an image, you might write:

"A glass perfume bottle on a stone pedestal, soft rim lighting."

For video, that's incomplete.

You need to explain what happens after the clip begins. Does the camera move toward the bottle? Do water droplets slide across the glass? Does the lighting change? Does mist move behind the object?

A prompt like "make a beautiful, dynamic commercial video" still leaves too many decisions to the model.

It might rotate the camera, move the object, or introduce unexpected changes in perspective.

The AI Video Prompt Formula

A practical formula is:

Scene → Subject → Action over time → Camera movement → Pacing and mood → Duration and format.

Example:

A transparent, unbranded perfume bottle stands on a dark stone pedestal. The scene begins with a medium shot of the bottle. The camera then slowly moves closer while small water droplets slide down the glass surface. A thin layer of mist drifts gently behind the product. Smooth continuous camera movement, no abrupt cuts, cool rim lighting, premium commercial product photography, horizontal 16:9 composition.

For text-to-video generation, the model creates both the scene and the action.

For image-to-video generation, the starting image already provides the visual composition. Your prompt should focus primarily on animation and movement.

Neuromia supports both text-to-video generation and creating video from an image used as the first frame.

If you're deciding which model to use, check the AI video model comparison to explore different approaches to characters, camera movements, and complex animation.

How to Describe Camera Movement in AI Video Prompts

Understanding a few basic filmmaking terms makes video prompting significantly easier.

Camera movement	Prompt wording	Best suited for
Dolly in	Camera slowly moves toward the subject	Products, faces, important details
Orbit	Camera moves in an arc around the subject	Showing the shape and volume of an object
Tracking shot	Camera follows a moving subject	Walking scenes, vehicles, characters
Static shot	Camera remains still while the subject moves	Food, nature, product animation
Handheld	Subtle natural camera shake	Documentary-style scenes, walking sequences

Avoid combining every camera movement into one short clip.

If the camera simultaneously pushes forward, circles the subject, tilts upward, and turns sharply, the resulting motion may become inconsistent.

Describe the subject's action as a sequence, too.

Instead of:

A woman dancing while the camera moves beautifully.

Try:

An adult woman wearing a long blue dress stands in a bright studio. She slowly raises her arms, then turns gently around her own axis as the fabric of her dress follows the movement. The camera stays at chest height and slowly moves to the right. One continuous shot, soft natural daylight.

You don't have to describe every movement of the hands or head.

Focus on the main sequence and avoid competing actions that make the scene harder to interpret.

Image-to-Video Prompts: What to Write When the First Frame Already Exists

When animating an existing image, you don't need to describe everything from scratch.

The model can already see the subject, clothing, background, and composition.

For example, if you upload an illustration of a fictional woman, try:

The woman slowly turns her head slightly to the right, blinks naturally, and gives a subtle smile. A gentle breeze creates minimal movement in her hair. The camera remains completely still. Preserve her original appearance, clothing, and composition. Smooth natural movements with no sudden facial changes.

For a product photograph:

The camera slowly moves closer to the product. Soft light reflections move gently across its surface. The product itself remains stationary, preserving its original shape and proportions. No additional objects or abrupt transitions.

By reducing unnecessary changes, you make it easier to evaluate whether the model handled the main animation correctly.

9 Ready-to-Use AI Video Prompts

These video generation prompts focus on short, self-contained scenes.

They can be used as starting points for text-to-video generation, while some instructions are also suitable for animating an existing image.

Set the duration and aspect ratio according to the options supported by your selected model. Including a format in the text does not necessarily change the generator's technical settings.

1. Nature: Ocean Waves at Sunrise

Cinematic, photorealistic shot of an empty sandy beach in the early morning. Small, clear waves slowly roll onto the shore and retreat, leaving a thin layer of water on the wet sand. The camera remains stationary, positioned low above the beach. Soft pink and blue sunrise on the horizon, gentle breeze, calm and natural water movement. One continuous shot, slow pacing, horizontal 16:9 format, approximately 5 seconds.

What to change: Weather, camera height, wave size, and time of day.

2. City: Rainy Street at Night

Photorealistic cinematic scene of a narrow city street in the evening after rainfall. Wet cobblestones reflect warm streetlights, while a few distant pedestrian silhouettes move slowly through the background. The camera glides forward along the sidewalk at eye level. Light atmospheric mist, realistic reflections, relaxed pacing, one continuous shot without cuts, horizontal 16:9 format.

What to change: Architecture, time of day, camera speed, and number of pedestrians.

3. Product Commercial: Rotating Perfume Bottle

Premium product commercial featuring a transparent glass perfume bottle filled with amber liquid, without branding, text, or logos. The bottle rotates slowly around its vertical axis on a glossy black pedestal. The camera remains stationary while directional lighting creates controlled moving reflections across the glass. Dark background, soft rim lighting, smooth continuous rotation, highly detailed product shape, horizontal 16:9 composition.

What to change: Product type, background color, rotation speed, and lighting setup.

4. Fantasy Character Walking Through a Forest

An original fictional traveler wearing a long green cloak walks along a narrow path through a fantasy forest filled with enormous glowing mushrooms. The camera tracks smoothly beside the character, maintaining a consistent distance. The cloak moves naturally with each step, while small glowing particles drift through the air. Soft blue-green lighting, magical atmosphere, relaxed pacing, one continuous shot, horizontal 16:9 format.

What to change: Character, environment, lighting colors, and movement direction.

5. Food Commercial: Pouring Coffee

Photorealistic close-up of a white ceramic cup resting on a wooden table. Fresh black coffee pours slowly into the cup in a steady stream from above. Small bubbles form on the surface while gentle steam rises. The camera remains completely still, focused on the cup and moving liquid. Warm morning light from a window on the left, natural reflections, calm commercial-style photography, no cuts, vertical 9:16 format.

What to change: Beverage, cup, background, and camera angle. Keeping one primary action can make the scene easier to control.

6. Animating a Fictional Character Portrait

Preserve the original appearance, facial features, hairstyle, and clothing of the illustrated woman. The character blinks naturally, slowly turns her head a few degrees to the left, and develops a very subtle smile. A light breeze gently moves a few strands of hair. The camera remains completely stationary, the background stays unchanged, and the animation is smooth without sudden distortions of facial proportions.

What to change: Gaze direction, facial expression, and amount of hair movement.

This prompt is designed for image-to-video generation using an existing illustration of a fictional character.

7. Mountain Cloud Timelapse

Static wide shot of a mountain valley viewed from a high hillside. In accelerated time, clouds move quickly across the peaks while shadows glide smoothly over the mountain slopes. The daylight gradually shifts toward evening. Mountains and trees preserve their shapes and positions. No camera movement, realistic atmospheric perspective, smooth timelapse without abrupt visual jumps, horizontal 16:9 format.

What to change: Landscape, time of day, and cloud conditions.

8. Abstract Motion Graphics Intro

Three translucent glass rings with soft blue illumination emerge against a completely dark background. The rings slowly approach one another, rotate smoothly, and form a symmetrical composition in the center of the frame. The camera moves slightly closer while subtle reflections and soft glowing highlights appear around the shapes. Clean contemporary 3D motion graphics, fluid animation, no text, logos, or additional objects, horizontal 16:9 format.

What to change: Geometric shapes, materials, colors, and movement direction.

9. Interior Walkthrough: Modern Living Room

Photorealistic modern living room with a large window, light wooden flooring, graphite cabinets, and a muted plum-colored sofa. The camera slowly moves from left to right, revealing the arrangement of furniture and depth of the room. Soft daylight enters through the window, creating natural shadows. Calm, controlled movement without architectural distortion. All furniture remains stationary, one continuous shot, horizontal 16:9 format.

What to change: Interior style, furniture, camera direction, and lighting.

Negative Prompts: What They Are and When to Use Them

A negative prompt describes elements you don't want to appear in the generated result.

These might include unwanted objects, text, excessive blur, or particular visual effects.

A regular prompt describes the desired scene:

A white ceramic cup on a wooden table, soft daylight, realistic product photography.

A negative prompt might specify:

Extra cups, unwanted objects, text, logos, heavy blur, distorted geometry.

Not every model provides a separate negative prompt field.

If the generator doesn't offer one, you can try including important restrictions directly in the main prompt:

"One ceramic cup, clean background, no text or logos."

Avoid adding dozens of restrictions automatically.

Negative prompts don't guarantee that every unwanted element will disappear, and too many competing constraints can make the task more complicated.

Start with a clear positive description. Add negative instructions when you notice a specific recurring problem.

What Language Should You Write AI Prompts In?

AI prompts can be written in English or other languages.

Most modern models understand many languages, particularly when the scene is straightforward and the instructions are unambiguous.

However, English terminology often provides the most predictable results for complex photography and filmmaking instructions.

Terms such as close-up, tracking shot, shallow depth of field, dolly in, and soft rim lighting are widely used to describe specific visual techniques.

In Neuromia, your prompt is sent to the selected model exactly as written, without automatic translation.

If you write your prompt in Spanish, French, German, or another language, the model receives that original text.

A practical approach is to start in the language you're most comfortable with. If the model struggles with a complicated camera movement or lighting setup, try an English version while keeping the scene itself unchanged.

For example:

Plain English:

Close-up shot of a red apple on a wooden table. Soft lighting from the left, blurred background, static camera.

More technical English:

Photorealistic close-up of a red apple resting on a rustic wooden table. Soft diffused key light from camera-left, shallow depth of field, subtle background bokeh, locked-off camera, natural color grading.

Both describe approximately the same scene, but the second version communicates more precise photographic choices.

You can experiment with different formulations using Nano Banana and other image models available in Neuromia.

When comparing prompts in different languages, try to preserve the subject, composition, and lighting so the comparison remains meaningful.

Maximum Prompt Length by AI Model

Different models support different maximum prompt lengths.

These are technical character limits, not recommended targets. A longer prompt is not automatically a better prompt.

Image generation models

Model	Maximum characters
Z-Image	1,000
Seedream 4.5 / 5.0	3,000
Flux Kontext	3,000
Nano Banana	5,000
Ideogram V3	5,000
Flux 2	5,000
Nano Banana Pro	10,000
GPT Image 2 / 2.5	20,000
Nano Banana 2	20,000

Video generation models

Model	Maximum characters
Kling 2.6	1,000
Hailuo Pro / Standard	1,500
Runway Gen-4 / Aleph	1,800
Kling 3.0	2,500
Seedance 2.0	20,000

Even if a model supports 20,000 characters, there's rarely a reason to fill the entire limit for a simple portrait or short video.

A few clear sentences may be enough.

When shortening a prompt, remove repeated adjectives and unnecessary background information first. Keep the subject, primary action, composition, and essential restrictions.

How Many Credits Does AI Generation Cost?

Neuromia uses credits for image and video generation.

The cost depends on the model and selected generation settings.

Here are example prices for generating one image at base quality:

Image model	Credits per image
Z-Image	0.88
Ideogram V3	3.85
Nano Banana 2.1	4.4
Flux 2 Pro	5.5
GPT Image 2.5	6.6
Seedream 5.0 Pro	7.7
Nano Banana 2	8.8
Midjourney	12.11
Nano Banana Pro	19.8

Video generation pricing varies by model and available options:

Video model	Starting price
Seedance 1.5 Pro	1.93 credits per second
Kling 3.0	15.4 credits per second
Runway Gen-4	13.2 credits per video
Hailuo 2.3 Standard	33 credits per video
Veo 3.1 Fast	66 credits per video

These are starting prices. The final cost may vary depending on supported generation settings, such as quality and duration.

New Neuromia users receive a 10-credit welcome bonus, which can be used toward available generations within the selected model's credit requirements.

For early experiments, start with a straightforward scene. Check whether the model understands the main composition before attempting more complex movements or detailed environments.

Don't Want to Write Prompts Yourself?

You don't need to build every AI prompt from scratch.

Neuromia offers several ways to turn a short idea into a more detailed instruction or start with an existing creative scenario.

The "Image and Video Prompts" Assistant

Neuromia includes an "Image and Video Prompts" assistant in the chat assistants section.

You provide a short idea, specify whether you're generating an image or video, and describe your preferred style.

For example:

I want a premium commercial video of a glass perfume bottle against a dark background. The camera should slowly move closer, with elegant product lighting.

The assistant creates a detailed English prompt, provides a Russian translation, and generates two alternative versions.

This is useful when you know what you want visually but aren't sure how to describe the camera angle, lighting, or animation.

You can also use it as a starting point for experimenting with different visual approaches.

Review the generated prompt before submitting it. Make sure the assistant hasn't added unnecessary actions, materials, or details that conflict with your original idea.

Ready-Made AI Templates

Another option is Neuromia's collection of ready-made templates.

These include scenarios for creating portraits from photos, restoring images, designing greeting cards, and other creative tasks.

The prompt is already prepared. Depending on the template, you only need to upload an image or select an available option.

Templates are particularly useful for repeatable tasks where you care more about the type of result than manually adjusting every detail.

The image editor also includes a "Prompt Examples" section. Select an example to insert its text into the prompt field, then customize it for your project.

Common AI Prompt Mistakes and How to Fix Them

Even detailed prompts don't always produce consistent results.

Often, the problem isn't the length of the prompt. It's conflicting instructions, vague descriptions, or an unclear main action.

Common mistake	How to fix it
"Make it beautiful and realistic"	Specify the subject, lighting, background, camera angle, and visual details
Too many objects in one image	Choose one main subject and make other elements secondary
Conflicting camera instructions	Use one clear camera angle or primary movement
Too many visual styles	Choose one direction and add two or three specific characteristics
No movement described in a video prompt	Explain what happens at the beginning, middle, and end
Rewriting the entire prompt after every attempt	Change one element at a time and compare results
Expecting flawless typography in generated images	Use short text, check the lettering, or add typography separately

Another frequent mistake is assuming that every disappointing result requires a longer prompt.

If the product isn't centered, you don't need another paragraph about texture or detail.

Just write:

A single product positioned precisely in the center, camera directly in front, symmetrical composition.

If the animation looks chaotic, reduce the number of actions and keep the camera stationary.

A useful workflow is:

Write the prompt → Generate → Identify the main problem → Adjust one instruction → Generate again.

Over time, you can build a personal collection of prompts that consistently communicate your preferred compositions, lighting setups, and types of movement.

FAQ: AI Prompts for Images and Videos

How Do You Write a Good AI Prompt?

Start with the main subject and action, then add the environment, visual style, lighting, and composition. For images, specify the camera angle and intended format. For videos, describe how the subject and camera move over time. Five clear instructions are often more useful than twenty vague adjectives.

What Is an AI Prompt in Simple Terms?

An AI prompt is a written instruction describing what you want an AI model to create. It might specify an image, video, or another type of output. The model uses your description to generate a result, so clear visual details help communicate your idea.

How Long Should an AI Prompt Be?

There's no universal ideal length. A simple product photo might need just a few sentences, while a complex scene may require a more detailed description. Always check the selected model's character limit, and prioritize useful instructions over unnecessary length.

Can AI Generate a Prompt for Me?

Yes. An AI prompt generator can turn a short idea into a detailed description. Neuromia's "Image and Video Prompts" assistant creates an English prompt, provides a Russian translation, and suggests two alternatives. Review the result to make sure the subject, composition, and action match what you actually want.

What Is a Negative Prompt, and Do I Need One?

A negative prompt specifies elements you want to avoid, such as unwanted objects, text, or excessive blur. It's optional, and not every model supports a dedicated negative prompt field. You can also include important restrictions directly in your main description.

What Language Is Best for AI Image and Video Prompts?

Start with the language that lets you describe your idea most clearly. Many modern models understand multiple languages, but English camera and lighting terminology can be more predictable for complex scenes. Neuromia passes your prompt to the selected model as written, without automatically translating it.

Final Thoughts

A good AI prompt isn't about using complicated terminology. It's about describing the result clearly enough for the model to understand what matters.

For images, focus on the subject, environment, lighting, and composition. For videos, add actions over time and camera movement.

Start with a simple prompt, change one element at a time, and save the versions that work best for your projects.

Choose any ready-to-use example from this guide, customize it for your idea, and try it in Neuromia's image or video generator. Practical experimentation is the fastest way to discover which instructions help you achieve the results you're looking for.