C.W.K.
Stream
VIDEO

Veo 3.1

Google DeepMind

Cinematic video generation with native synchronized audio — dialogue, SFX, and music from one prompt.

Veo 3.1 Complete Prompting Guide

A Human-Friendly Reference for Video Generation with Native Audio


Welcome: What Is Veo 3.1 and Why Should You Care?

Imagine describing a scene — the creak of a detective's chair, rain pattering against a tin roof, a woman saying three quiet words that change everything — and watching it appear as a short, cinematic video clip, complete with sound, in under three minutes. That's what Veo 3.1 makes possible.

Veo 3.1 is Google DeepMind's flagship video generation model. Its defining feature isn't just visual quality — it's native audio: dialogue, sound effects, ambient noise, and music generated in sync with the video, from a single prompt. No separate audio tracks. No post-production sound design needed. The model reads your description and conjures both picture and sound simultaneously.

Released in October 2025 with audio and editing capabilities, then upgraded with 4K resolution support in January 2026, Veo 3.1 represents a significant leap beyond its predecessor Veo 3, which was largely silent or had very limited audio. Google Cloud Blog

Why Native Audio Matters

Before native audio, AI video workflows looked like this: generate video → record or license audio → sync manually → export. It was multi-tool, multi-skill, and often jarring in the final result because audio and visuals were generated independently with no shared understanding of the scene.

With Veo 3.1, the model "knows" what the scene sounds like at the same time it renders it. A glass shattering on a marble floor in frame zero plays at exactly the right moment. A character's lips move in sync with their dialogue. Footsteps on gravel match the walker's pace. This synchronization is where Veo 3.1 genuinely changes the game — especially for solo creators, small studios, and anyone producing content without a full production team.

The Two Variants: Fast vs. Quality

Veo 3.1 comes in two flavors, and understanding when to use each is one of the highest-leverage decisions you'll make as a user.

Veo 3.1 Fast (veo-3.1-fast-generate-001) is roughly 2.2× faster than Quality and costs about half as much. It produces excellent results — 8.7/10 overall quality in independent testing — and is nearly indistinguishable from Quality mode for most social media and web use cases. Use it constantly during the creative process.

Veo 3.1 Quality (veo-3.1-generate-001) is the premium tier: deeper detail, smoother motion, more accurate lighting. It scores 9.2/10 overall and is what you want for client-facing deliverables, brand commercials, or anything that will be projected on a large screen. Use it for finals.

The practical workflow: explore and iterate in Fast, render finals in Quality. Apiyi.com


Quick Start: Three Prompts You Can Use Right Now

Not ready to read the whole guide? Here are three copy-paste prompts that demonstrate Veo 3.1's capabilities. Each one follows the best-practice formula and will produce a usable result immediately.

Prompt 1 — Simple (No Audio Focus)

PROMPT
Duration 6s. AR 16:9. Fast mode.

Wide shot of a golden retriever running across a sun-drenched meadow at magic hour, 
wildflowers in the foreground slightly blurred by shallow depth of field. 
Camera gently tracks alongside the dog at ground level. 
Style: warm, cinematic, vivid color.

Prompt 2 — With Audio Direction

PROMPT
Duration 8s. AR 9:16. Fast mode.

Medium close-up on a 35mm lens: a barista in her late 20s with dark curly hair 
concentrating as she pours steamed milk into a latte, creating a leaf pattern. 
Camera is locked, slightly low angle looking up at her hands.

Audio: the hiss of the steam wand, the gentle clink of the cup on the counter, 
soft café background chatter, acoustic guitar playing quietly from a speaker. 
She glances up at camera and says, "Enjoy," with a warm smile. No subtitles.

Prompt 3 — Timestamp Prompting (Multi-Shot)

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:02] Extreme wide shot: a lone lighthouse on rocky cliffs at dusk, 
waves crashing far below. Camera is static, horizon level.

[00:02-00:05] Slow dolly-in on the lighthouse beam rotating, golden light sweeping 
across the ocean surface. SFX: wind, crashing waves.

[00:05-00:08] Interior close-up: a weathered lighthouse keeper's hand turns a 
brass dial. He says in a gravelly voice, "Clear skies by morning." 
Ambient: the hum of the lighthouse mechanism, sea wind muffled through glass.

Part 1: Technical Specifications

Model Identity

FieldValue
Product NameVeo 3.1
DeveloperGoogle DeepMind
Initial ReleaseOctober 2025 (audio + editing); January 2026 (4K update)
CategoryVideo generation with native audio
VariantsVeo 3.1 Quality (Standard), Veo 3.1 Fast
AccessGoogle AI Studio, Vertex AI, LTX Studio, invideo, Visla, and other third-party platforms
PredecessorVeo 3 (silent/limited audio)

Core Technical Specs (Both Variants)

ParameterValue
Resolutions720p, 1080p, 4K (3840×2160) — 4K added January 2026
Aspect Ratios16:9 (horizontal), 9:16 (vertical)
Clip Durations4, 6, or 8 seconds per generation
AudioNative synchronized audio (dialogue, SFX, ambient, music)
Reference ImagesUp to 3 reference images or 1 style image ("Ingredients to Video")
Start/End FrameSupported (with audio)
Add/Remove ObjectSupported (currently uses Veo 2 model, no audio)
Video Extension+7 seconds per extension, up to 20 extensions
WatermarkingSynthID on all generated videos

Fast vs. Quality: Full Comparison

DimensionVeo 3.1 FastVeo 3.1 Quality (Standard)
Model IDveo-3.1-fast-generate-001veo-3.1-generate-001
Generation Time (8s video)~1 min 13s~2 min 41s
Speed Multiplier2.2× fasterBaseline
Overall Quality Score8.7/109.2/10
Detail Richness8.3/109.0/10
Motion Smoothness8.9/109.3/10
Lighting Realism8.5/109.1/10
Audio Sync9.0/109.1/10
Cost per 8s video~$0.10–$0.15~$0.20–$0.75
Best ForRapid iteration, social media, batch productionFinal delivery, cinematic, high-end brand

The quality gap between Fast and Quality is often difficult to distinguish with the naked eye. Fast mainly lags in extreme detail rendering — complex textures, subtle lighting gradients. For most use cases, start with Fast for exploration, use Quality only for finals. Apiyi.com


Part 2: The Prompt Formula

The Five-Part Formula (Start Here)

Every effective Veo 3.1 prompt contains five elements, in this order:

PROMPT
[Cinematography] + [Subject] + [Action] + [Context] + [Style & Ambiance]
ElementDescriptionExamples
CinematographyCamera work and shot composition"Medium shot", "Crane shot ascending", "POV shot", "Tracking shot"
SubjectMain character or focal point with specific traits"A tired corporate worker in a wrinkled white shirt"
ActionWhat the subject is doing (concrete verbs)"rubbing his temples in exhaustion"
ContextEnvironment and background elements"in front of a bulky 1980s computer in a cluttered office late at night"
Style & AmbianceOverall aesthetic, mood, lighting"Retro aesthetic, shot as if on 1980s color film, slightly grainy"

Why this order matters: Veo 3.1 is trained on cinematographic conventions. Leading with the camera shot tells the model how to frame everything else. A "medium shot" of a subject immediately constrains composition, depth of field, and spatial relationships — before the model has even processed your subject description. Think of it as giving the model its instructions before handing it a script.

The Extended Seven-Element Formula (Maximum Control)

For complex, production-ready shots, expand to seven elements:

PROMPT
[Camera & Lens] + [Subject] + [Action & Physics] + [Environment] + [Lighting] + [Style & Texture] + [Audio]

The extra elements — lighting and audio — are where Veo 3.1's strengths really come through. Specifying lighting as a distinct element (rather than bundling it into "style") gives the model clearer, non-conflicting instructions. Specifying audio separately ensures all three audio layers (dialogue, SFX, ambient) get attention rather than the model guessing what the scene should sound like.

Example using all seven elements:

PROMPT
Duration 8s. AR 16:9. Quality mode.

[Camera & Lens] Low-angle tracking shot, 24mm wide-angle lens with slight barrel distortion.
[Subject] A street food vendor in her 50s, worn apron, weathered hands.
[Action & Physics] Expertly flipping thin crepes on a cast-iron griddle, batter sizzling on contact.
[Environment] A narrow Paris alley at dawn, cobblestones glistening from overnight rain.
[Lighting] Soft warm glow from a single overhead market lamp, cold blue dawn light 
creeping in from the alley entrance.
[Style & Texture] Film grain, slightly desaturated, documentary aesthetic.
[Audio] SFX: batter sizzle, spatula scraping metal. Ambient: distant pigeons, 
early morning city hum. No dialogue.

Part 3: Cinematography Language

The cinematography element is the most powerful single tool in Veo prompts. The model has deep understanding of professional film grammar — use it. LTX Studio

Camera Movements

MovementPrompt LanguageEmotional EffectWhat This Looks Like
Dolly"Dolly shot", "slow dolly-in"Intimacy, approachThe camera physically moves toward (or away from) the subject on a track or wheeled rig. Unlike a zoom, the background perspective actually changes — objects in the background shift relative to the subject. A slow dolly-in on a face feels like leaning in to listen to a secret.
Tracking"Tracking shot alongside"Following actionThe camera moves parallel to a moving subject — think of a runner filmed from a car keeping pace. It creates a sense of momentum and puts the viewer "in motion" alongside the subject.
Crane"Crane shot ascending"Scale revealThe camera rises on a mechanical arm from a low position to a high one. A crane shot that starts at ground level and pulls up to reveal a vast landscape is one of cinema's most powerful "reveal" moments — the world suddenly gets bigger.
Pan"Slow pan left to right"Surveying sceneThe camera rotates horizontally on a fixed axis (like turning your head). Used to survey a space or follow action moving across the frame. A slow pan across a battle aftermath feels methodical and mournful.
POV"POV shot"ImmersionThe camera literally represents a character's eyes. The viewer becomes the character. Used in everything from horror (the monster approaching from the victim's perspective) to action (the athlete's view from inside a car).
Handheld"Shaky handheld"Documentary, urgencyThe camera is held by a human operator and moves with natural, slight instability. This imperfection signals "reality" — news footage, documentary, or gritty realism.
Steadicam"Smooth Steadicam tracking"Polished followA gyroscopically stabilized handheld rig that eliminates shake while still moving through space. The result feels fluid and slightly ethereal — gliding through environments rather than cutting between them. Think of the famous Kubrick hotel corridor shots.
Static/Locked"Camera locked at eye level"Stability, authorityThe camera doesn't move at all. This lets action within the frame do all the work. A locked-off medium shot of a speaker feels authoritative and direct — nothing distracts from what they're saying.

Shot Sizes (Composition)

ShotPrompt LanguageShowsWhat This Looks Like
Extreme Close-Up (ECU)"Extreme close-up"Single detail (eye, hand, texture)Fills the entire frame with one small detail — an eye, a fingertip, a crack in concrete. Used for intense emotional moments or to reveal detail invisible at normal scale. The world disappears; only this one thing exists.
Close-Up (CU)"Close-up"Face or detail, emotional intimacyShows just the face (roughly chin to forehead) or a hand-sized object. The go-to shot for emotion — you can read micro-expressions. The audience feels like they're in a private moment with the subject.
Medium Close-Up (MCU)"Medium close-up"Chest upFrom chest to top of head. The standard TV and interview shot — close enough to feel engaged, wide enough to see some body language and gesture.
Medium Shot (MS)"Medium shot"Waist up, conversationalFrom waist to top of head. The "normal" conversational distance — how we see people when talking face to face. Neither too intimate nor too distant.
Wide Shot (WS)"Wide shot"Full scene, contextShows the full subject within their environment. The subject is clearly visible but the setting is now equally important. Used to establish "where are we?"
Extreme Wide (EWS)"Extreme wide shot"Vast landscape, subject tinyThe subject may be barely visible against a massive environment. Used to convey scale, isolation, or the smallness of human beings against nature or architecture.
Two-Shot"Two-shot"Two characters in frameBoth characters share the frame — usually a medium or wide framing. Establishes relationship and spatial proximity. The viewer can watch the interaction without cutting back and forth.
Low Angle"Low angle shot"Power, dominanceCamera is below the subject, pointing upward. The subject looms, appears powerful, even threatening. Politicians and heroes are often filmed this way.
High Angle"High-angle shot"VulnerabilityCamera is above the subject, pointing downward. The subject appears smaller, more vulnerable, observed. Can feel like surveillance or god's-eye perspective.

Lens and Focus Choices

TechniquePrompt LanguageWhat This Looks Like
Shallow DOF"Shallow depth of field, f/1.8"The subject is sharp; everything behind (and in front) melts into soft blur. This is the "bokeh" look that immediately signals a high-end camera. It isolates the subject and creates a dream-like quality.
Wide-angle"Wide-angle lens, slight barrel distortion"A wider field of view captures more of the scene. Objects at the edges of the frame curve slightly inward (barrel distortion). Makes spaces feel larger and more expansive. Action looks more dynamic.
Macro"Macro lens"Extreme close-focus capability that reveals microscopic details — the surface of a strawberry, water droplets on a watch face, the weave of a fabric. Makes the tiny world monumental.
Deep focus"Deep focus, everything sharp"Both foreground and background are in focus simultaneously. Used deliberately to make every element in the frame equally important — or to create a slightly unsettling, hyper-real effect.
Soft focus"Soft focus, dreamy"A subtle diffusion of the entire image. Nothing is razor sharp. Romantic, nostalgic, or dreamlike — the visual equivalent of looking through gauze.
Anamorphic"Anamorphic lens, horizontal flares"The widest, most cinematic look available. Produces a distinctive oval bokeh, compressed depth, and horizontal lens flares that streak elegantly across bright light sources. This is the "blockbuster movie" look.

Part 4: Audio Direction

Veo 3.1's native audio is its headline feature. The key to great audio results is treating the soundstage in three distinct layers — and being as specific about sound as you are about visuals. invideo

Layer 1: Dialogue

Use quotation marks for any specific spoken lines. Include a description of the speaker and their emotional delivery.

PROMPT
A woman in her 30s looks directly at camera and says, "We have to leave now."

Best practices for dialogue:

  • Specify voice quality: "in a weary voice", "whispers", "shouts excitedly", "with a British accent", "voice cracking with emotion"
  • Add "No subtitles" if you don't want auto-generated text overlays
  • For conversations, describe who speaks, in what order, and their emotional state
  • Avoid writing more than 2–3 sentences of dialogue per 8-second clip — it rushes the delivery

Layer 2: Sound Effects (SFX)

Describe sounds with specificity. Use the SFX: prefix to make the audio instruction unambiguous.

PROMPT
SFX: Thunder cracks in the distance, followed by the patter of rain on a tin roof.

Common SFX patterns:

  • "SFX: footsteps on gravel", "SFX: glass shattering", "SFX: car engine revving"
  • "The door creaks open slowly" (integrating SFX naturally into the scene description)
  • "A coffee cup clinks against the saucer"
  • The more specific the surface and material, the more realistic the result: "hard-soled shoes on marble" vs. just "footsteps"

Layer 3: Ambient Noise

Define the background soundscape — the "room tone" of the scene. This is often the difference between a video that feels alive and one that feels like a rendering.

PROMPT
Ambient noise: the quiet hum of a starship bridge, occasional beeping of instruments.
PROMPT
Audio: Gentle café ambience — quiet conversation, espresso machine hiss, light jazz in background.

Layer 4: Music

PROMPT
A swelling, gentle orchestral score begins to play.
PROMPT
Audio: lo-fi hip hop beat, soft piano, vinyl crackle.

Audio Scene Examples by Genre

Dialogue Scene (Drama)

PROMPT
Duration 8s. AR 16:9. Quality mode.

Two-shot in a warmly lit kitchen. A mother (50s) and daughter (20s) stand on 
opposite sides of an island counter.

Dialogue: Mother (quietly, not making eye contact): "I just want you to be happy." 
Daughter (voice catching): "Then let me make my own choices."

SFX: The hum of the refrigerator. A tea kettle begins to whistle faintly in the background.
Ambient: Quiet suburban evening, distant cricket sounds.

ASMR

PROMPT
Duration 8s. AR 9:16. Quality mode.

Macro lens, extreme close-up on two hands slowly unwrapping a textured handmade 
soap bar, revealing layers of lavender and oat. Camera is perfectly still.

Audio: The soft crinkle of paper wrapping, the quiet scrape of fingernails across 
soap texture, no music. Absolute quiet beyond the hands.

Action Sequence

PROMPT
Duration 6s. AR 16:9. Fast mode.

Handheld camera, low angle tracking shot: a motorcycle rider in a leather jacket 
leans hard into a tight turn on a rain-slicked mountain road. Tire spray arcs behind.

SFX: Engine roar peaking, tires squealing on wet asphalt, wind whipping past the 
microphone. No music — pure mechanical sound.

Ambient Nature

PROMPT
Duration 8s. AR 16:9. Quality mode.

Extremely wide static shot: a dense redwood forest at dawn. Light rays cut through 
morning mist between massive trunks. Nothing moves except drifting mist.

Audio: Complete forest quiet — distant woodpecker, single crow call, wind in high 
branches, absolute stillness beneath. No music. No human sounds.

Comedy

PROMPT
Duration 6s. AR 16:9. Fast mode.

Medium shot: a golden retriever sits perfectly still at a dining table, wearing 
a tiny bow tie, while a man off-camera tries to get its attention. 
The dog stares at the camera with total composure.

Dialogue (man, off-camera, increasingly desperate): "Biscuit. Biscuit. Look at me. 
BISCUIT."
SFX: Biscuit sound effect from off-screen. Dog does not move.
Audio: A gentle comedic tuba note. Audience laughter track optional.

Horror

PROMPT
Duration 8s. AR 16:9. Quality mode.

Slow dolly-in through the darkness toward a half-open basement door. 
A sliver of yellow light from inside. No subject visible.

SFX: The groan of a settling house, the creak of a floorboard from below the door, 
the distant drip of water. Ambient: near-total silence with subwoofer drone 
slowly building. No music — just the drone.

Part 5: Negative Prompting

Use positive exclusion language rather than negative constructions. Veo 3.1 responds better to descriptions of what IS in the scene than to what ISN'T. Visla Blog

Instead ofUse
"No man-made structures""A desolate landscape with no buildings or roads"
"No people""An empty, abandoned space"
"No modern elements""Period-accurate, pre-industrial setting"
"No noise""Complete silence — no audio"
"Not a happy scene""A somber, grief-stricken atmosphere"

For specific exclusions, append a clean Exclude: line at the end:

PROMPT
Exclude: logos, extra text, crowds, modern vehicles, lens flare.

Part 6: Advanced Workflows

Workflow 1: Timestamp Prompting (Multi-Shot in Single Generation)

Timestamp prompting lets you direct a complete multi-shot sequence within a single 8-second generation. Instead of generating multiple clips and editing them together, you write a mini shot list embedded in the prompt itself. LTX Studio

The format:

PROMPT
[00:00-00:02] First shot description.
[00:02-00:05] Second shot description.
[00:05-00:08] Third shot description.

From the official guide:

PROMPT
[00:00-00:02] Medium shot from behind a young female explorer with a leather satchel, 
pushing aside jungle vines to reveal a hidden path.
[00:02-00:04] Reverse shot of her freckled face, expression filled with awe, 
ancient moss-covered ruins visible in the background. 
SFX: rustling leaves, distant exotic bird calls.
[00:04-00:06] Tracking shot following her as she runs her hand over intricate carvings 
on a crumbling stone wall. Emotion: wonder and reverence.
[00:06-00:08] Wide, high-angle crane shot revealing the lone explorer standing small 
in the center of a vast forgotten temple complex. 
SFX: swelling gentle orchestral score begins.

Workflow 2: Image-to-Video

When animating a source image, the rules change significantly:

  • Do NOT redescribe what's visible in the image — the model already has that information
  • Focus your prompt entirely on action, camera movement, and audio
  • The image defines subject, setting, composition, and style
PROMPT
The subject slowly turns their head toward camera and smiles. 
Gentle push-in on a 50mm lens. 
Audio: ambient room tone, soft fabric rustling.

Workflow 3: First and Last Frame (Transition Prompting)

Provide a start frame image and an end frame image. Your prompt describes the camera path and transformation between them.

PROMPT
The camera performs a smooth 180-degree arc shot, starting with the front-facing 
view of the singer and circling around her to end on the POV shot from behind her 
on stage. She sings "when you look me in the eyes, I can see a million stars." 
Crowd cheers build in volume as the camera rotates.

Workflow 4: Ingredients to Video (Character and Style Consistency)

Provide up to 3 reference images plus 1 style image. The model uses these for consistent characters, objects, or visual styles across multiple shots — critical for any project requiring continuity.

PROMPT
Using the provided images for the detective, the woman, and the office setting: 
create a medium shot of the detective behind his desk. He looks up at the woman 
and says in a weary voice, "Of all the offices in this town, you had to walk into mine."

Workflow 5: Video Extension

Extend existing clips by +7 seconds per extension, for up to 20 extensions. Useful for building longer sequences from a strong initial generation.


Part 7: Timestamp Prompting Masterclass

Timestamp prompting is one of Veo 3.1's most powerful and underutilized capabilities. The following three sequences demonstrate how to write full multi-shot scenes for different production contexts.


Sequence 1: Product Launch Commercial (30-Second Equivalent via Extension)

This example uses 8 seconds of timestamp prompting to establish a brand story opening. Subsequent extensions would continue the narrative.

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:02] Extreme close-up, macro lens: the tip of a pen touching paper. 
Ink flows into the first letter of a signature. 
SFX: the quiet scratch of pen on quality paper. Ambient: total silence.

[00:02-00:05] Pull back to medium shot: we see the signer — a woman in her 40s 
in a crisp charcoal blazer — closing a leather portfolio. 
She sets the pen down with precision. 
SFX: portfolio latch clicking shut. Ambient: high-floor office quiet.

[00:05-00:08] Wide shot from behind her left shoulder: floor-to-ceiling windows 
reveal a rain-swept city skyline. She stands and looks out. 
Camera slowly pushes toward the window over her shoulder.
Audio: A single, resonant piano note. 
She says quietly: "Let's begin."

Sequence 2: Nature Documentary Opener

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:03] Extreme wide, aerial drone shot descending slowly: an Arctic tundra 
at blue hour, completely flat, dusted with snow. A single polar bear walks 
across the white expanse from right to left. She is very small in the frame.
Ambient: Arctic wind, low and constant.

[00:03-00:06] Cut to medium tracking shot at ice level, camera tracking alongside 
the bear as she walks. Her breath steams. Her paws press into snow with each step.
SFX: the soft crunch of each footfall in frozen snow. Wind continues.

[00:06-00:08] Slow dolly-in on her face in close-up. She pauses, lifts her head, 
scans the horizon. Eyes alert.
Ambient: Wind drops to near silence. A held breath. 
Narrator (warm, quiet voice): "She hasn't eaten in eleven days."

Sequence 3: Short Film Opening Scene (Dialogue-Driven)

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:02] Wide establishing shot: a 1970s-era diner, near empty, fluorescent 
buzz, rain against the windows. A man sits alone in a corner booth.
Ambient: rain on glass, distant highway sounds, diner AC hum.

[00:02-00:05] Medium shot across the table — the man's perspective looking toward 
the entrance. A woman in a wet coat pushes through the door, scanning the room.
SFX: diner door chime, rain burst as it opens and closes.

[00:05-00:08] Two-shot as she slides into the seat across from him. 
Both stare at each other for a beat.
Man (low, flat): "You're late."
Woman (unwinding scarf, not apologizing): "I had to make sure I wasn't followed."
Ambient: the diner continues its quiet background life. The rain intensifies.

Part 8: The Essential Rules

These rules, applied consistently, will produce dramatically better results. Google Cloud Blog

  1. Lead with camera. "Wide aerial", "Medium handheld", "Macro product shot" — always first. This is the most impactful single change you can make to any prompt.
  2. Use concrete verbs. "Opens the umbrella" instead of "experiences a rainy moment". Veo 3.1 translates verbs into motion; vague verbs produce vague motion.
  3. One subject per shot. Add a secondary subject only when they interact. Competing subjects confuse the model's attention.
  4. One dominant action per clip. Multiple conflicting actions create motion instability. If your character needs to stand up, walk across the room, AND open a door — those are three separate clips.
  5. State duration and aspect ratio. "Duration 6s; AR 9:16; Audio on". Always declare these at the top of your prompt.
  6. Specify audio layer by layer. Dialogue, SFX, ambient — each as a distinct instruction. Mixing them into a single audio sentence often results in the model prioritizing one and ignoring others.
  7. Describe what you see and hear, not how to build it. Tell Veo what the camera captures, not the technical pipeline. You're writing a scene description, not a production spec.

Part 9: Production-Ready Example Prompts

Cinematic Product Reveal

Original example (from the base guide):

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:03] Open on a dark, minimalist studio setup with a premium wireless 
headphone in soft silhouette on a matte surface. Camera stays steady, building 
anticipation. Ambient: low electronic hum.
[00:03-00:06] Slow side-light sweep as the camera gently pushes closer, revealing 
form, texture, and brushed aluminum details. SFX: soft mechanical click.
[00:06-00:08] Clean close-up, headphone fully lit, camera settles confidently. 
The text overlay fades in. Audio: subtle bass-heavy electronic note resolves.

Variation 2 — Skincare Product:

PROMPT
Duration 6s. AR 9:16. Quality mode.

[00:00-00:02] Overhead flat-lay shot: a white marble surface. A single glass 
serum bottle placed from above by elegant hands. Slow push-in from directly above.
SFX: the soft click of glass on marble.

[00:02-00:06] Camera arcs slowly around the bottle at low angle. Morning light 
catches the amber serum inside, glowing. Fine details: pipette texture, 
minimal label typography. 
Audio: Single sustained singing bowl tone, fading in. No dialogue. No music.
Exclude: price text, website URLs, logos within video.

Variation 3 — Watch Reveal:

PROMPT
Duration 8s. AR 16:9. Quality mode.

Macro lens, rotating turntable shot: a luxury mechanical watch face slowly 
rotates counterclockwise. The movement visible through a sapphire crystal caseback.
Lighting: dramatic single-source sidelight with a dark background.

SFX: The faint ticking of the movement. Ambient: near-total silence.
Style: Fine detail, deep blacks, jewel-toned reflections.
Camera slowly pulls back as the watch face rotates to reveal the full case.

Variation 4 — Food and Beverage:

PROMPT
Duration 6s. AR 9:16. Fast mode.

Medium close-up: a bartender's hands drop a large clear ice sphere into a crystal 
glass with a single confident motion. Dark amber whiskey pours in a smooth arc 
over the ice. The ice cracks faintly as cold spirits hit it.

SFX: The sharp crack of ice, the gentle glug of the pour, 
the faint sizzle of condensation on glass.
Lighting: warm backlight through the amber liquid.
Style: Cinematic, rich contrast, slow motion (50% speed).

Talking Head (Brand/UGC)

Original example:

PROMPT
Duration 6s. AR 9:16.

Camera locked at eye level, medium close-up on a 35mm lens: a startup founder 
in his late 30s with short black hair and light stubble, wearing a charcoal cotton 
hoodie, speaking directly to camera. He leans slightly forward, lifts one hand to 
emphasize a point, then relaxes back to neutral. Quiet office during late afternoon, 
blurred monitors glowing faintly in the background. Lit by soft daylight from a 
side window with gentle fill on the opposite side. Style: fine skin pores, 
visible fabric weave, subtle contrast, no gloss.

Audio: The founder says, "This update cuts setup time in half, helping teams get 
started faster." No subtitles. Soft office ambience.

Variation 2 — Wellness/Lifestyle Creator:

PROMPT
Duration 6s. AR 9:16. Fast mode.

Medium close-up, camera locked, eye level: a woman in her late 20s with natural 
hair and minimal makeup, wearing a soft white linen top, speaking conversationally 
to camera. She's seated in front of a large bookshelf, warm morning light from 
the right. She makes a small, open-palmed gesture mid-sentence.

Audio: She says with a relaxed, confident tone: 
"I stopped tracking every single calorie and honestly? I feel so much better." 
No subtitles. Ambient: birdsong from an open window.

Variation 3 — Corporate/B2B:

PROMPT
Duration 8s. AR 16:9. Quality mode.

Medium shot, slight low angle (camera just below eye level), locked: 
a woman in her 40s in a tailored navy blazer, silver earrings, professional 
but warm. She stands against a clean, blurred boardroom background. 
Confident, deliberate delivery.

Audio: "Our clients see an average of 34% faster close rates in the first quarter. 
That's not a projection — that's what the data shows." 
SFX: None. Ambient: quiet corporate HVAC hum. No subtitles.

Variation 4 — Comedy/Entertainment Creator:

PROMPT
Duration 6s. AR 9:16. Fast mode.

Medium close-up, handheld (slight natural movement): a man in his early 30s 
with an incredulous expression, gesturing expressively. He's in a home kitchen. 
Casual, slightly chaotic energy.

Audio: He says in an exasperated-but-amused voice: 
"I followed the recipe EXACTLY. Exactly. Tell me why it looks like this." 
SFX: A dramatic comedic sting. Ambient: light kitchen sounds.

Nature / Documentary

Original example:

PROMPT
Duration 8s. AR 16:9.

Crane shot starting low on a lone hiker and ascending high above, revealing they 
are standing on the edge of a colossal, mist-filled canyon at sunrise. 
Epic fantasy style, awe-inspiring, soft morning light. 
The hiker's jacket flutters in the wind. Camera continues ascending until 
the hiker is a small figure against the vast landscape.

Audio: Wind howling softly, distant eagle cry, swelling orchestral strings.

Variation 2 — Underwater:

PROMPT
Duration 8s. AR 16:9. Quality mode.

Wide underwater shot, looking up toward the surface: a blue whale moves 
slowly overhead from left to right, its vast shadow darkening the water. 
Shafts of sunlight filter down from above, bending around the whale's form.

Audio: Low, resonant whale song. Deep underwater ambience — bubbles, 
the muffled world above the surface, deep water pressure hum. No music.

Variation 3 — Time of Day Transition:

PROMPT
Duration 8s. AR 16:9. Quality mode.

Extremely wide, static tripod shot: a wheat field just before and during 
the moment of sunrise. Time appears to move at 10x speed. 
Shadows race across the field as the sun clears the horizon. 
Color shifts from deep blue-grey to warm amber gold.

Audio: Wind through wheat field — a constant, shifting rush. 
Birds beginning to call as light increases. No music.

Variation 4 — Intimate Macro Nature:

PROMPT
Duration 6s. AR 16:9. Quality mode.

Macro lens, extreme close-up: a single dew drop on a spider web strand. 
The drop contains a miniature reflection of the entire garden behind the camera. 
Camera is perfectly still. A light breeze causes the strand to vibrate almost 
imperceptibly.

SFX: Near silence — the distant sound of wind in trees, a single bee passing. 
Ambient: a quiet morning in a garden. Style: natural colors, no enhancement.

Social Media Vertical

Original example:

PROMPT
Duration 4s. AR 9:16. Fast mode.

Close-up of a hand placing a perfectly frosted cupcake on a marble counter. 
Confetti sprinkles cascade in slow motion. Warm, soft overhead lighting. 
Pastel color palette. Clean, satisfying ASMR aesthetic.

Audio: Gentle tap as cupcake touches marble, soft sprinkle sounds, no music.

Variation 2 — Fitness/Gym:

PROMPT
Duration 6s. AR 9:16. Fast mode.

Low-angle tracking shot at knee height: heavy-duty sneakers hit a treadmill belt, 
pace building quickly. The camera slowly rises to reveal the runner — a woman in 
her 30s, focused, earbuds in. Black gym background, neon accent lighting.

SFX: Treadmill motor hum increasing, rhythmic footfall. 
Audio: High-energy electronic music fades in under the footsteps.
Style: High contrast, slightly desaturated except for neon.

Variation 3 — Travel Hook:

PROMPT
Duration 4s. AR 9:16. Fast mode.

POV shot: a hand reaches forward to part two hanging curtains, 
revealing a balcony with a view of the Santorini caldera at golden hour. 
Sea is glittering. A glass of wine sits on the railing.

SFX: Fabric parting, ocean breeze, distant waves. 
Audio: A single bright piano note. No dialogue.

Variation 4 — Book/Reading Aesthetic:

PROMPT
Duration 4s. AR 9:16. Fast mode.

Overhead flat-lay, slow dolly-in: an open book on a white linen surface, 
a cup of tea steaming beside it, a pressed flower as a bookmark. 
Autumn leaves visible in soft focus through a window.

Audio: The sound of rain on glass, the quiet rustle of a page turning, 
the soft clink of the tea cup settling. No music. No dialogue.
Style: Desaturated, warm ivory and brown tones.

Multi-Character Dialogue

Original example:

PROMPT
Duration 8s. AR 16:9. Quality mode.

Using the provided reference images for Character A (detective) and Character B (witness):

Medium two-shot in a dimly lit interrogation room. The detective leans forward 
across a metal table. The witness sits with arms crossed, looking away. 
A single overhead fluorescent light flickers.

Dialogue:
Detective (gruff voice): "You were the last person to see him alive."
Witness (nervous, quiet): "I already told the other officer everything."

SFX: fluorescent light buzzing, chair creaks. 
Ambient: muffled sounds from beyond the door.

Variation 2 — Romance:

PROMPT
Duration 8s. AR 16:9. Quality mode.

Medium two-shot, handheld, slightly shaky: two people stand in a rainy parking 
lot under a single streetlight. Both are getting wet. Neither moves first.

Dialogue:
Her (steady, measured): "Tell me why you didn't say anything."
Him (quietly, looking down): "Because I didn't think you'd stay."

SFX: Rain on asphalt, a distant car passing, the sound of water dripping 
from a sign overhead.

Variation 3 — Comedy Scene:

PROMPT
Duration 8s. AR 16:9. Fast mode.

Medium two-shot, locked camera: a competitive cooking show set. 
Two contestants stare at an enormous pile of mystery ingredients. 
Both look equally horrified.

Dialogue:
Contestant A (whispering without moving lips): "Is that a durian?"
Contestant B (also not moving): "Don't look at it. Just don't look at it."

SFX: Dramatic cooking show music sting, then silence.
Ambient: Kitchen sounds, distant judges murmuring.

Variation 4 — Historical Drama:

PROMPT
Duration 8s. AR 16:9. Quality mode.

Medium two-shot with elegant period staging: a Victorian drawing room, 
candlelight. A lord stands by the fireplace; a woman in period dress 
sits perfectly upright in a wingback chair. Tension.

Dialogue:
Lord (controlled, cold): "You cannot be serious about this arrangement."
Woman (with absolute composure): "I have never been more serious about anything."

SFX: Fire crackling, a clock ticking. 
Ambient: The creak of the house settling in winter cold.

Brand Commercial

Variation 1 — Outdoor/Adventure Brand:

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:03] Extreme wide drone shot: a group of four hikers cresting a ridgeline 
at dawn, silhouetted against a burning orange sky. Anamorphic lens, 
wide horizontal flares from the sun.
Audio: Silence, then a single guitar chord fades in.

[00:03-00:08] Low-angle wide shot from the ground in front of them: 
the group stops at the top, one of them raises both arms. 
Camera slowly pushes upward as they cheer.
Voiceover (calm, confident): "Built for every summit."
SFX: Wind, the sound of boots on rocky ground. Music builds.

Variation 2 — Tech Product:

PROMPT
Duration 6s. AR 16:9. Quality mode.

Medium shot, clean white studio: a person's hand opens a premium laptop. 
The screen powers on and illuminates their face with soft light.
They begin typing — fast, confident, in flow.

Audio: The clean sound of laptop keys, a subtle startup chime, 
then an upbeat minimal electronic track at low volume.
Voiceover: "Your work. At full speed."
Style: Minimal, clean, white and silver palette.

Variation 3 — Food and Hospitality:

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:04] Wide establishing shot: a beautiful outdoor restaurant terrace 
at sunset overlooking the Mediterranean. Tables set with white linen. 
Candles. Guests laughing softly.
Ambient: Soft laughter, glasses clinking, gentle sea breeze.

[00:04-00:08] Camera moves to a medium close-up on a plate being set down — 
fresh seafood, beautifully plated, steaming slightly.
SFX: The plate touching the table, a light sizzle from the dish.
Voiceover: "This is what a good day tastes like."

Tutorial / Explainer

Variation 1 — Software Tutorial:

PROMPT
Duration 8s. AR 16:9. Fast mode.

Screen-capture style medium shot over someone's shoulder: a designer's hands 
on a keyboard, screen showing a creative software interface.
She selects a layer, applies a blend mode, and the image transforms dramatically.

Audio: She narrates: "Click the layer, set the blend to Screen, 
and watch the background disappear." 
SFX: Mouse clicks, keyboard tap. Ambient: quiet studio. No music.
Style: Clean and professional, screen clearly legible.

Variation 2 — Cooking Tutorial:

PROMPT
Duration 6s. AR 9:16. Fast mode.

Overhead flat-lay, medium close-up: a cutting board. 
An experienced cook's hands dice an onion in fast, confident strokes.

Audio: She narrates: "Keep the root end intact — it holds everything 
together while you cut." 
SFX: Knife on wooden board, the clean thwack of each cut. 
Ambient: Kitchen sounds. No music.

Variation 3 — Fitness Tutorial:

PROMPT
Duration 8s. AR 16:9. Fast mode.

Wide side-angle shot, clean gym space: a trainer demonstrates a deadlift.
He performs a rep, then pauses at the top.

Audio: He narrates: "Hinge at the hips, not the lower back. 
Feel the tension in your hamstrings before you pull."
SFX: The bar touching the ground, a breath out at the top.
Style: Clean gym, no distractions. Arrows or text overlays for form cues — 
add in post.

Part 10: Common Mistakes → Better Version

Mistake 1: Vague Visual Description

Before:

A dramatic product reveal of our new headphones.

After:

PROMPT
Duration 8s. AR 16:9. Quality mode.
[00:00-00:03] Macro lens, extreme close-up on the hinge mechanism of a 
folded headphone — dark studio, single rim light. Camera slowly orbits.
[00:03-00:06] Camera pulls back to reveal the full headphone silhouette 
as a side-light sweeps left to right, revealing brushed aluminum texture.
SFX: A soft mechanical click. Ambient: Low electronic hum.

Mistake 2: Ignoring Audio Layers

Before:

A barista making coffee with nice sound.

After:

PROMPT
Duration 6s. AR 9:16.
Close-up of an espresso portafilter being locked into the machine. 
The barista tamps the grounds, locks the filter, and pulls the shot — 
a dark crema streams into a white ceramic cup.
SFX: Tamp thud, metal click of the portafilter, the rushing hiss of the espresso pull.
Ambient: Café background — quiet chatter, jazz, the bark of a coffee grinder elsewhere.
No dialogue. No music overlay.

Mistake 3: Multiple Conflicting Actions

Before:

PROMPT
A woman walks into a room, looks around, finds what she's looking for, 
picks it up, smiles, and walks back out.

After (three separate generations):

Clip 1:

PROMPT
Duration 4s. Medium shot: A woman pushes open a door and steps into 
a dimly lit study, pausing just inside the threshold.

Clip 2:

PROMPT
Duration 4s. Medium close-up: Her eyes scan the room, then fix on 
something off-camera to the right.

Clip 3:

PROMPT
Duration 4s. Close-up on her hands: She picks up an old key from a 
dusty desk. Her expression softens — relief.

Mistake 4: Using Negative Language

Before:

PROMPT
No buildings, no cars, no modern elements. Not a happy scene.

After:

PROMPT
A desolate pre-industrial landscape — barren fields, a distant 
treeline, a grey winter sky. The atmosphere is heavy and mournful.
Exclude: modern structures, vehicles, technology.

Mistake 5: No Camera Direction

Before:

A scientist discovering a cure, dramatic moment.

After:

PROMPT
Duration 6s. AR 16:9. Quality mode.
Low-angle, slow dolly-in: a scientist in her 60s — face illuminated 
by a laptop screen in a dark lab — stares at her results. 
Her expression shifts from disbelief to wonder. 
She exhales slowly, sets down her pen.
SFX: The quiet hum of laboratory equipment. 
No dialogue. No music. Let the silence speak.

Mistake 6: Forgetting to Specify Format

Before:

PROMPT
A beautiful sunrise over mountains, ethereal and cinematic.

After:

PROMPT
Duration 8s. AR 16:9. Quality mode.
Extremely wide static shot: jagged mountain peaks at first light. 
The sky transitions from deep violet to amber gold over 8 seconds. 
A thin layer of cloud hugs the valleys, catching the first light.
Ambient: Wind, a single distant bird call. No music.
Style: Documentary realism, natural color, no enhancement.

Mistake 7: Redescribing Source Image in Image-to-Video Mode

Before:

PROMPT
[Uploading a photo of a cat sitting on a windowsill]
Prompt: "A fluffy orange tabby cat sitting on a windowsill, 
looking out at the rain."

After:

PROMPT
[Uploading the same photo]
Prompt: The cat's tail slowly swishes once. 
It turns its head slightly toward camera. 
Ambient: rain on glass, the quiet hum of a house.

Mistake 8: Expecting Long-Form Video from One Generation

Before:

PROMPT
Create a 60-second documentary about ocean conservation.

After (Approach): Plan eight 8-second clips using timestamp prompting. Generate each clip, then stitch:

PROMPT
Clip 1 (Duration 8s, AR 16:9, Quality):
[00:00-00:04] Wide underwater shot: vibrant coral reef, schools of fish.
[00:04-00:08] Cut to bleached, dead coral. Stark silence.
Narrator: "In 50 years, we've lost half the world's coral."

[Continue with clips 2-8, each with its own timestamp structure]

Use Video Extension to add +7s to strong clips before stitching.


Part 11: Fast vs. Quality Decision Guide

A practical flowchart for deciding which model to use:

PROMPT
START: What is this video for?

├── Is this an exploratory or test generation?
│   └── YES → USE FAST. Always iterate in Fast.
│
├── Is this for social media (Reels, TikTok, Shorts, Stories)?
│   └── YES → USE FAST. The compression at those platforms 
│             will hide any quality difference.
│
├── Is this one of many clips in a batch production run?
│   └── YES → USE FAST. Cost savings are significant at scale.
│
├── Will this be viewed on a large screen or projected?
│   └── YES → USE QUALITY. Detail rendering matters at scale.
│
├── Is this a final deliverable for a paying client?
│   ├── If client work for web/social → Evaluate Fast first.
│   └── If client work for broadcast/cinema/large format → USE QUALITY.
│
├── Is budget the primary constraint right now?
│   └── YES → USE FAST. Quality at Fast's price point is 
│             genuinely excellent (8.7/10).
│
├── Does the shot have extreme texture detail you need to preserve?
│   (fine fabric weave, complex natural surfaces, 
│    hair in motion, micro-expressions)
│   └── YES → USE QUALITY. This is where Fast most visibly lags.
│
└── Default rule:
    USE FAST for the first 3-5 attempts on any new prompt.
    SWITCH TO QUALITY only for the final approved version.

The bottom line: The 0.5 overall quality gap (8.7 vs. 9.2) between Fast and Quality is real but narrow. You will pay 2-5× more for Quality. In most production pipelines, spending that budget on more iterations in Fast will yield better final results than fewer iterations in Quality. Apiyi.com


Part 12: Prompt Building Workshop

This section walks through building a complex prompt from scratch — the way a professional would actually approach it.

The Brief

Goal: A 8-second cinematic product video for a premium fountain pen brand. Tone: heritage, craftsmanship, quiet luxury. Platform: brand website hero video (horizontal), and a 9:16 crop for Instagram Reels.


Step 1: Define Your Format Controls

Start by declaring your technical parameters before writing any description:

PROMPT
Duration 8s. AR 16:9. Quality mode.

(You'll generate a second version at AR 9:16 for mobile.)


Step 2: Choose Your Camera Story

What does the camera do in this clip? A product hero video needs to feel like a reveal. Think: what does this pen deserve?

Option A: Static macro, the pen just sits there perfectly. (Safe, clean.) Option B: Slow orbit around the pen on a turntable. (Dynamic, shows all angles.) Option C: Timestamp prompting — start with extreme close-up detail, pull back to the full pen. (Best storytelling arc for 8 seconds.)

Choose C. Write the camera structure first:

PROMPT
[00:00-00:03] Extreme close-up, macro lens...
[00:03-00:07] Camera slowly pulls back...
[00:07-00:08] Wide enough to see the full pen...

Step 3: Add Subject and Physical Detail

Now describe the pen specifically — material, color, era, condition:

PROMPT
[00:00-00:03] Extreme close-up, macro lens: the nib of a vintage-style 
gold fountain pen. 18-karat two-tone nib, deeply engraved tipping.

Step 4: Add Action and Physics

What happens to the pen? It should do something — not just sit:

PROMPT
...The nib lightly touches the surface of cream-colored writing paper. 
A thin line of midnight blue ink flows from the nib tip — 
we watch the ink settle into the paper fibers.

Step 5: Add Environment and Lighting

Where is this? What does light do?

PROMPT
...Set on a dark walnut writing desk. Lighting: a single warm desk lamp 
from camera left, casting a pool of gold light. Deep shadows surround.

Step 6: Add Style and Texture

What film stock, what aesthetic? This is premium heritage — think warm, rich, analog:

PROMPT
Style: Cinematic, deep shadow, rich amber tones. 
Slight film grain. The aesthetic of a well-made watch advertisement.

Step 7: Write the Audio

Heritage luxury is quiet. Sound design should feel intimate:

PROMPT
SFX: The near-silent whisper of nib on fine paper. 
Ambient: A quiet room — a clock ticking very softly in the background. 
No music. Let the silence carry weight.

The Complete Prompt

PROMPT
Duration 8s. AR 16:9. Quality mode.

[00:00-00:03] Extreme close-up, macro lens: the nib of a vintage-style gold 
fountain pen — 18-karat two-tone nib, deeply engraved tipping — lightly 
touches cream-colored writing paper. A thin line of midnight blue ink flows 
from the nib tip; we watch it settle into the paper fibers.

[00:03-00:07] Camera slowly pulls back, revealing the full body of the pen: 
deep burgundy lacquer, gold trim, a cap resting beside it. The pen rests 
on a dark walnut writing desk. Lighting: a single warm desk lamp from 
camera left, casting a pool of gold. Deep shadows surround.

[00:07-00:08] Camera settles into a composed medium close-up. 
The pen lies perfectly still. The ink line on the paper completes.

SFX: The near-silent whisper of nib on fine paper.
Ambient: A quiet room — a clock ticking very softly in the background.
No music. No dialogue. Let the silence carry weight.

Style: Cinematic, deep shadow, rich amber tones. Slight film grain. 
Watch advertisement aesthetic.

Exclude: price labels, modern elements, plastic surfaces, harsh lighting.

Step 8: Iterate

Generate this in Fast first. Check:

  • Is the macro close-up sharp on the nib?
  • Does the ink flow look realistic?
  • Does the pullback feel smooth?
  • Is the audio appropriately quiet and luxurious?

Adjust any element that doesn't work. Once the prompt is refined and you're happy with the composition and motion, run Quality for the final hero asset.


Part 13: Format Controls — Quick Reference

ControlOptions
Duration4s, 6s, 8s
Aspect Ratio16:9 (horizontal), 9:16 (vertical)
Resolution720p, 1080p, 4K
AudioOn (default) or specify "No audio"
Extension+7s increments, up to 20×
ModelFast (veo-3.1-fast-generate-001) or Quality (veo-3.1-generate-001)

Cheat Sheet: Veo 3.1 at a Glance

The Core Formula

PROMPT
[Camera] + [Subject] + [Action] + [Environment] + [Style] + [Audio]

Always Lead With Camera

Dolly / Tracking / Crane / Pan / POV / Handheld / Steadicam / Static

Shot Size Controls Intimacy

ECU → CU → MCU → MS → WS → EWS (intimate to epic)

Low Angle = Power | High Angle = Vulnerability

Audio: Three Layers, Always

Dialogue: "Quote in quotation marks."
SFX: Specific sounds, specific surfaces.
Ambient: The room/world background sound.

Timestamp Format

[00:00-00:02] Shot one.
[00:02-00:05] Shot two.
[00:05-00:08] Shot three.

Fast vs. Quality Decision

Exploring? → FAST
Social media? → FAST
Batch production? → FAST
Client final, broadcast, large format? → QUALITY

Format Controls Header (Always Include)

PROMPT
Duration [4s/6s/8s]. AR [16:9/9:16]. [Fast/Quality] mode.

Top 5 Rules

  1. Camera first, always
  2. Concrete verbs ("pours" not "experiences")
  3. One subject, one dominant action per clip
  4. Positive framing, not negative ("empty street" not "no cars")
  5. Image-to-video: describe motion only, not what's already visible

Max Clip Length

8 seconds per generation. Use Video Extension (+7s) or stitch clips for longer content.

Common Lens Choices

  • "Anamorphic" → blockbuster cinematic
  • "Macro lens" → intimate product/nature detail
  • "Wide-angle, slight barrel distortion" → expansive, dynamic
  • "Shallow depth of field, f/1.8" → subject isolation, premium look

Sources: Google Cloud Blog, WaveSpeed AI, Apiyi.com, Visla Blog, LTX Studio, invideo, PXZ AI. Guide compiled and expanded from the original machine-readable prompting guide. Last updated: March 2026.

Spotted a bug or have feedback on this page?Report an Issue

Comments 0

🔔 Reply notifications (sign in)
Sign inPlease sign in to comment.

No comments yet — be the first.