
Shown in Higgsfield throughout the course: https://higgsfield.ai/
Seedance creates amazing videos. This means your prompt should focus on shots, movement, actions, and storytelling rather than describing a single image/visual
.
One of the most effective prompting methods in Seedance is to think like a filmmaker. Instead of writing one paragraph describing a scene, break your video into individual shots and describe what happens in each one.
Always start by specifying:
For example:
Total: 15 seconds / 6 shots / 16:9
This helps Seedance understand the structure of your video before it begins generating.
A strong Seedance prompt typically follows this structure:
Describe the overall look of the video.
Example:
Describe an overview of the whole video scene you want, including:
Example:
"A young filmmaker sits alone on a rooftop at sunset overlooking the city. He reviews old footage on a laptop while storm clouds gather in the distance."
Describe each shot individually.
Example:
Shot 1: Wide shot establishing the rooftop at sunset.
Shot 2: Medium shot of the filmmaker reviewing footage.
Shot 3: Close-up of the laptop screen showing childhood memories.
Shot 4: Camera slowly pushes in as he becomes emotional.
Seedance performs best when movement is clearly described.
Describe:
Examples:
The best videos have a clear sequence of events.
A common structure is:
Normal Situation → Conflict → Action → Resolution
Example:
This progression helps Seedance generate more cinematic results.
Total: 12 seconds / 5 shots / 16:9
A teenage boy (describe) stands alone in a stormy field (describe) at night. Lightning flashes across the sky.
Shot 1: Wide shot of the boy standing in the field.
Shot 2: Lightning strikes nearby, illuminating his face.
Shot 3: Close-up as energy begins flowing through his body.
Shot 4: Medium shot as he transforms into a glowing superhero.
Shot 5: Wide shot of the transformed hero standing confidently as the storm clears.
Total: 10 seconds / 4 shots / 16:9
A rancher works on a cattle farm at sunrise.
Shot 1: Drone shot revealing the landscape.
Shot 2: Medium shot of the rancher opening a gate.
Shot 3: Tracking shot alongside moving cattle.
Shot 4: Wide shot of the sunrise over the ranch.
Total: 8 seconds / 3 shots / 16:9
Luxury wristwatch advertisement.
Shot 1: Close-up of the watch face.
Shot 2: Slow camera orbit around the watch.
Shot 3: Wide hero shot with dramatic studio lighting.
❌ Describing only a static image.
"A man standing on a rooftop."
✅ Describe actions and shots.
"A man stands on a rooftop while the camera slowly pushes toward him as city lights appear behind him."
❌ Using only one camera angle.
✅ Use multiple shots and camera movements.
❌ Forgetting to specify duration and shot count.
✅ Always include:
❌ Trying to tell an entire movie in one prompt.
✅ Focus on one short sequence with a clear beginning, middle, and end.
These terms can help Seedance avoid artificial or game-like visuals and produce more realistic results.

Unlike some AI video generators that favour long, highly detailed prompts, Runway Gen-4 performs best when you start simple. The goal is to describe the motion you want to see and then refine the result by gradually adding more detail.
Runway recommends beginning with a basic prompt that focuses on the most important movement in the scene. Once that works well, you can add camera movement, scene movement, and style details to improve the result.
Think like a filmmaker directing motion rather than describing a static image.
A strong Runway prompt typically follows this structure:
Subject Motion + Camera Motion + Scene Motion + Style
Start simple:
"A cowboy rides through a desert."
Then refine:
"A cowboy rides through a desert as a handheld camera follows from behind. Dust trails behind the horse. Cinematic western style."
What is the main subject doing?
Examples:
Example:
"A teenage girl walks along abandoned train tracks."
How is the camera moving?
Examples:
Example:
"A handheld camera tracks beside her."
What is moving in the environment?
Examples:
Example:
"Wind blows through tall grass while dust drifts across the scene."
How should the video look?
Examples:
Example:
"Cinematic live-action with warm golden-hour lighting."
"A teenage girl walks along train tracks at sunset. A handheld camera tracks beside her. Dust moves through the air. Cinematic live-action."
"A rancher rides across open countryside. A drone camera slowly pulls back. Cattle move through the field. Documentary style."
"A luxury watch rotates slowly on a black surface. The camera orbits around the product. Light reflections move across the metal. Premium commercial style."
"A futuristic motorcycle speeds through a neon city. The camera follows closely behind. Rain falls and reflects city lights. Cinematic cyberpunk style."
❌ Starting with an overly complex prompt.
✅ Begin simple and add details gradually.
❌ Describing only the subject.
"A man standing on a rooftop."
✅ Include motion.
"A man stands on a rooftop while a slow camera push-in reveals the city skyline behind him."
❌ Forgetting environmental movement.
✅ Add scene motion.
"Leaves blow through the street as the character walks forward."
❌ Trying to perfect the prompt immediately.
✅ Iterate one element at a time.
Add:
One step at a time.
Remember: Runway Gen-4 is designed for iteration. Start simple, test often, and build complexity step by step.

Googles Video on Flow: https://www.youtube.com/watch?v=oKjDeMtBZ4g
Veo 3.1 is designed to understand filmmaking language. The best prompts describe not only what happens in the scene, but also how it is filmed, how it sounds, and how it feels.
Google recommends using a structured prompt that combines camera direction, subject, action, environment, and visual style. Think like a film director, cinematographer, and sound designer all at once.
The more clearly you define these elements, the more cinematic and controlled your video will be.
Google recommends this structure:
[Cinematography] + [Subject] + [Action] + [Context] + [Style & Ambiance]
"Medium shot, a tired office worker rubbing his temples, sitting at a cluttered desk late at night. The room is illuminated by fluorescent lights and the glow of an old computer monitor. Retro 1980s film aesthetic, slightly grainy, melancholic mood."
How is the scene filmed?
Examples:
Example:
"Close-up with shallow depth of field."
Who or what is the focus of the scene?
Examples:
Example:
"A young filmmaker holding an old camera."
What is happening?
Examples:
Example:
"Walking slowly along abandoned train tracks."
Where is the scene taking place?
Examples:
Example:
"At sunset with a dusty town visible in the distance."
What should the scene feel like?
Examples:
Example:
"Cinematic realism, warm golden-hour lighting, reflective mood."
"Tracking shot, a teenage girl walking along abandoned train tracks, looking toward the horizon, with a dusty southern town in the background. Cinematic realism, golden-hour lighting, emotional atmosphere."
"Drone shot, a rancher riding horseback across open countryside, herding cattle through rolling hills at sunrise. Natural lighting, documentary style, peaceful mood."
"Close-up shot, a luxury wristwatch rotating slowly on black marble, reflections moving across the polished metal surface. Premium commercial aesthetic, studio lighting, shallow depth of field."
"Wide shot, an astronaut stepping onto the surface of a distant planet, examining an alien structure beneath towering mountains. Cinematic science fiction, dramatic lighting, epic scale."
One of Veo's biggest strengths is its ability to generate audio along with video.
Use quotation marks for spoken lines.
Example:
"A woman looks toward the storm and says, 'We have to leave now.'" And, how does she say it (confidently, angrily, shy, with an [insert] accent).
Describe important sounds.
Example:
"SFX: thunder cracks in the distance."
Describe the background soundscape.
Examples:
Example:
"Ambient sound: gentle ocean waves and distant seagulls."
❌ Describing only the subject.
"A man standing on a rooftop."
✅ Include cinematography, action, and style.
"Wide shot of a man standing on a rooftop overlooking the city at sunset, wind moving through his jacket, cinematic lighting."
❌ Forgetting camera language.
✅ Use filmmaking terms.
❌ Ignoring sound.
✅ Include:
When relevant.
❌ Using vague instructions.
"Make it dramatic."
✅ Describe what creates the drama.
"Dark storm clouds gather overhead while distant thunder echoes across the landscape."
Remember: Veo 3.1 is built for storytelling. Combining strong cinematography, action, sound, and atmosphere will usually produce the most cinematic results.

Kling works best when prompts clearly describe the subject, movement, environment, camera language, lighting, and atmosphere. Unlike some video generators that focus heavily on cinematography or shot sequences, Kling's strength lies in creating a clear visual scene from a concise, structured description.
Think of your prompt as describing a short 5–10 second movie clip. Focus on what the viewer sees, what the subject is doing, and how the scene feels.
Kling supports:
Kling recommends the following structure:
Subject + Subject Movement + Scene + Camera Language + Lighting + Atmosphere
"A young filmmaker holding a vintage camera, walking slowly through a deserted street. Small-town main street at sunset. Medium shot with shallow depth of field. Golden-hour lighting. Cinematic and nostalgic atmosphere."
Who or what is the focus of the video?
Examples:
Example:
"A teenage girl with long dark hair wearing a denim jacket."
What is the subject doing?
Examples:
Example:
"Walking along abandoned train tracks."
Where does the action take place?
Examples:
Example:
"An old railway line outside a dusty southern town."
How is the scene filmed?
Examples:
Example:
"Medium shot with a blurred background."
What type of lighting is used?
Examples:
Example:
"Warm sunset lighting with lens flare."
What should the scene feel like?
Examples:
Example:
"Cinematic and reflective mood."
"A teenage girl with windswept hair walks slowly along train tracks. A small southern town sits in the distance. Medium shot, shallow depth of field, golden-hour lighting, cinematic and nostalgic atmosphere."
"A rancher rides horseback across rolling countryside. Sunrise over open fields. Aerial shot, natural morning light, peaceful documentary atmosphere."
"A luxury wristwatch slowly rotating on a black marble surface. Close-up shot with blurred background. Studio lighting, premium commercial atmosphere."
"A futuristic motorcycle races through a neon-lit city. Skyscrapers glow in the background. Low-angle shot, neon lighting, cyberpunk atmosphere."
"A giant panda wearing black-rimmed glasses reads a book in a café. A steaming coffee sits beside the book near a window. Medium shot, ambient lighting, cinematic colour grading."
❌ Making prompts overly complicated.
✅ Use simple, visual descriptions.
❌ Describing too many actions in one short video.
✅ Focus on a single moment or action.
❌ Forgetting camera language.
✅ Add:
❌ Ignoring lighting and atmosphere.
✅ Include:
Remember: Kling performs best when you clearly describe a short visual moment rather than a long story. Focus on one scene, one action, and one mood for the strongest results.

Midjourney videos work differently from most AI video generators. Instead of creating an entire scene from text, you start with an image—either one you've created in Midjourney or uploaded yourself—and then animate it.
Because the image already contains the characters, environment, lighting, and composition, your prompt doesn't need long descriptions. Instead, focus on describing the movement you want to see.
Think of Midjourney Video as animating a photograph rather than generating a complete video from scratch.
A strong Midjourney video prompt typically follows this structure:
Subject Movement + Camera Movement + Motion Style
Unlike other video tools, you usually don't need to describe:
These are already contained in your starting image.
What is moving?
Examples:
Example:
"She slowly turns toward the camera."
How should the camera move?
Examples:
Example:
"Slow cinematic push-in."
How much movement should there be?
Midjourney offers:
Low Motion
High Motion
Starting Image:
Portrait of a young filmmaker.
Prompt:
"Slow push-in as she looks toward the horizon, hair moving gently in the wind."
Starting Image:
Cowboy standing in a desert.
Prompt:
"Dust blows across the landscape as the camera slowly circles around him."
Starting Image:
Floating castle above the clouds.
Prompt:
"Clouds drift past the castle as the camera glides forward."
Starting Image:
Luxury watch render.
Prompt:
"Slow camera orbit around the watch, subtle reflections moving across the metal surface."
❌ Re-describing the image.
"A woman wearing a red jacket standing in a forest at sunset..."
If this is already visible in the image, you don't need to repeat it.
✅ Focus on movement.
"She turns slowly toward camera as leaves drift past."
❌ Writing long cinematic paragraphs.
Midjourney usually performs better with shorter motion-focused prompts.
❌ Using High Motion for every video.
High Motion can create exciting results but may also introduce unrealistic movement.
✅ Start with Low Motion for more natural animations.
Best for:
Examples:
Best for:
Examples:
Remember: Midjourney video is less about describing a scene and more about bringing an existing image to life. The strongest prompts usually focus on simple, natural movement rather than complex storytelling.

Grok works best when prompts clearly describe the subject, action, setting, camera perspective, and mood. Unlike tools such as Seedance that often require detailed shot-by-shot instructions, Grok responds well to natural language descriptions and iterative refinement.
Think of Grok as a creative collaborator. Start with a simple scene description and then add camera, movement, and atmosphere details as needed. Grok can generate videos from text alone or animate existing images.
Subject + Action + Setting + Camera + Style/Mood
"A surfer carving through a giant wave at sunrise, wide-angle camera, cinematic lighting, slow-motion, epic atmosphere."
Who or what is the focus?
Examples:
What is happening?
Examples:
Where does the scene take place?
Examples:
How is the scene filmed?
Examples:
Examples:
"A lone cowboy rides through a desert at sunset, wide-angle tracking shot, dust blowing across the landscape, cinematic western style."
"A rancher moves cattle across open countryside at sunrise, drone shot, natural lighting, documentary atmosphere."
"An astronaut explores an abandoned space station, floating debris drifting through the corridors, cinematic science-fiction mood."
❌ Being too vague.
"Create a cool video."
✅ Be specific.
"A sports car racing through a neon-lit city at night, low-angle tracking shot, cyberpunk atmosphere."
❌ Overcomplicating prompts.
✅ Focus on one clear scene and action.
Meta's video generation tools work best with conversational prompts that clearly describe the scene, action, environment, and mood. You don't need complex formulas or highly technical language.
Think of explaining a scene to a filmmaker: who is there, what they're doing, where they are, and how the scene should feel.
Subject + Action + Environment + Style/Mood
"A young filmmaker walks through a busy city street at night, neon signs reflecting in puddles, cinematic atmosphere."
Who or what is the focus?
Examples:
What are they doing?
Examples:
Where does it happen?
Examples:
How should it feel?
Examples:
"A teenage girl walks along train tracks at sunset, warm golden light, emotional cinematic atmosphere."
"A traveller explores narrow streets in a historic Italian town, handheld camera feel, vibrant colours, documentary style."
"A floating city above the clouds at sunrise, epic fantasy atmosphere, cinematic lighting."
❌ Being too generic.
"Create a fantasy video."
✅ Add context.
"A floating castle above the clouds at dawn, golden sunlight, epic fantasy atmosphere."
❌ Adding too many unrelated ideas.
✅ Focus on one scene.
Remember: Meta AI generally responds best to clear scene descriptions rather than highly technical prompt engineering.

A good Suno prompt clearly describes the type of music you want to create. Rather than focusing on visual details or storytelling, focus on the song's genre, mood, instruments, vocals, and structure.
Suno can generate complete songs with lyrics or instrumental music. For the best results, describe the sound as if you were briefing a music producer.
If you want maximum control, use Custom Mode, where you can separately define lyrics, music style, and title.
A strong Suno prompt typically follows this structure:
Genre + Mood + Instruments + Vocals + Purpose/Theme
"Cinematic orchestral soundtrack, emotional and inspiring, piano and strings, building percussion, instrumental, suitable for a motivational short film."
What type of music do you want?
Examples:
Example:
"Cinematic orchestral soundtrack."
What emotion should the song create?
Examples:
Example:
"Inspirational and hopeful."
What instruments should be featured?
Examples:
Example:
"Piano, strings, and soft percussion."
What type of vocals do you want?
Examples:
Example:
"Female vocal with an emotional delivery."
What is the song for?
Examples:
Example:
"Music for a cinematic travel documentary."
"Cinematic orchestral soundtrack, uplifting and inspirational, piano, strings, and percussion, instrumental, suitable for a motivational short film."
"Lo-fi hip hop, relaxed mood, soft piano chords, vinyl crackle, mellow drums, instrumental, study and productivity music."
"Modern pop, upbeat and energetic, synths and electronic drums, female vocal, catchy chorus about chasing your dreams."
"Contemporary country, emotional and heartfelt, acoustic guitar and fiddle, male vocal, story about small-town life and perseverance."
"Epic cinematic trailer music, powerful brass, strings, massive percussion, instrumental, heroic and dramatic atmosphere."
For greater control, Suno allows you to provide:
Custom Mode is ideal when creating songs for films, YouTube videos, or branded content.
Example:
Style of Music
"Cinematic folk, acoustic guitar, warm strings, emotional, uplifting, male vocal."
Lyrics
[Verse]
I've been walking down this road...
[Chorus]
Every step brings me closer home...Using section tags such as [Verse], [Chorus], and [Bridge] can help Suno follow the song structure more effectively.
❌ Being too vague.
"Make a good song."
✅ Be specific.
"Upbeat indie pop, bright guitars, energetic drums, female vocal, feel-good summer anthem."
❌ Only specifying a genre.
"Rock song."
✅ Add mood, instruments, and vocal style.
"Alternative rock, emotional tone, electric guitar and drums, male vocal."
❌ Trying to describe every detail.
✅ Focus on the most important musical elements.
❌ Forgetting the intended use.
✅ Tell Suno what the music is for.
"Background music for a travel montage."
Remember: Suno performs best when you describe the sound and emotion of the music, not the visuals. Think like a songwriter, composer, or music producer when writing your prompts.
Genre + Mood + Instruments + Vocals + Purpose
Example:
"Epic cinematic orchestral, inspiring and emotional, piano and strings, instrumental, suitable for a motivational documentary trailer."
AI Video School
Copyright © 2026 AI Video School - All Rights Reserved.