AI Video Generation Prompts for Faceless Videos

A selfie-style vlog shot on a phone front camera, handheld at arm's length.

SUBJECT: [describe person: mid-20s woman, curly dark hair, olive green windbreaker].
LOCATION: [e.g. a windy clifftop path above the sea at golden hour].
ACTION: she walks toward camera while talking, wind moving her hair, occasionally glancing past the lens at the view.
CAMERA: front-facing phone lens, slight wide-angle distortion, natural handheld micro-shake, subject fills the left two-thirds of frame.
LIGHTING: warm low sun behind her creating a rim light, face lifted by soft sky bounce.
AUDIO: her voice close and slightly breathy from walking, real wind noise, distant waves, no music.
DIALOGUE (spoken by the subject): "[write one line, under 20 words, conversational, no narration voice]"

Photorealistic, natural colour, 8 seconds, no subtitles, no text overlays, no cuts.

First-person POV, camera mounted at eye level on a person walking.

SCENE: a dense pine forest at 2am. A handheld torch beam sweeps left to right across trunks and low mist. Breath fogs faintly at the bottom of frame.
MOVEMENT: slow forward walk, natural footstep bounce, the beam leading the eye, one hesitant pause when a sound comes from the right and the beam swings toward it.
CAMERA: 18mm wide, deep shadows, heavy natural vignetting, torch is the only light source, everything beyond 8 metres falls to black.
AUDIO: crunching pine needles underfoot, distant owl, one sharp branch snap off-screen right, breathing close to the mic. No music, no narration.
GRADE: desaturated cold blues in shadow, warm tungsten torch beam, visible sensor grain.

Photorealistic, continuous single take, 8 seconds, no text.

Extreme close-up macro video for ASMR.

SUBJECT: [e.g. hands slowly slicing a translucent bar of kinetic sand with a chrome blade].
CAMERA: 100mm macro lens, f/4, subject fills the entire frame, shallow focus with a slow rack focus across the action, locked-off tripod, zero camera shake.
MOTION: deliberate and slow, one continuous action from start to finish, no cuts.
LIGHTING: soft top light with a subtle warm rim, gentle specular highlights on every edge, clean dark background.
AUDIO: binaural, very close mic — crisp cutting sound, fine granular crumble, soft fabric contact, no music, no voice, no room echo.
FINISH: photorealistic, 4K detail, satisfying and tactile.

8 seconds, seamless loop-friendly start and end pose, no text.

Anime-style cooking sequence in the visual language of a modern Japanese food anime.

SCENE: [e.g. a small wooden kitchen at dawn, steam catching the light from a paper window].
ACTION: hands lift the lid of a clay pot, steam bursts upward in slow motion, broth ripples, a ladle draws up noodles that stretch and glisten.
STYLE: hand-drawn cel animation, clean line work, lush painted backgrounds, exaggerated food shine and steam volume, sakuga-quality motion on the hero action.
CAMERA: starts wide, pushes in fast on the pot, snap cut to macro on the broth surface, then slow pull back.
AUDIO: bubbling broth, sizzling, a soft chime, gentle piano underscore. No dialogue.
GRADE: warm ambers and soft golds, glowing highlights.

8 seconds, no subtitles, no text on screen.

Top-down time-lapse of a cardboard DIY build.

SETUP: overhead camera locked directly above a wooden workbench, frame filled edge to edge with the work surface.
ACTION: two hands cut, fold, glue and assemble [object: a working mini vending machine] from plain corrugated cardboard. The build progresses visibly: flat sheets, cut pieces, folded body, assembled structure, final working mechanism demonstrated once at the end.
CAMERA: static overhead, no zoom, no pan.
LIGHTING: even soft daylight, no harsh shadows across the work area.
AUDIO: crisp blade cuts, cardboard flex, tape peel, hands moving — no music, no voice.
FINISH: photorealistic, sharp texture on the corrugated edges, natural colour.

8 seconds, smooth accelerated motion, no text overlays.

Macro video of hand embroidery in progress.

SUBJECT: a needle pulling [colour] thread through taut natural linen in a wooden hoop, forming a [motif: single petal of a poppy].
CAMERA: 100mm macro, f/3.5, framed tight on the needle entry point, tiny slow dolly-in over the shot.
MOTION: one continuous stitch — needle enters, thread pulls through with slight fabric tension, thread settles.
LIGHTING: soft directional window light from the left, delicate shadow in the fabric weave, gentle sheen on the thread.
AUDIO: the faint pop of the needle through fabric, thread drag, ambient room tone only.
FINISH: photorealistic, visible individual linen fibres and thread twist, natural colour.

8 seconds, no text, no music.

Soft 3D animated lullaby scene for young children.

SCENE: a cosy round bedroom in a treehouse, a sleepy [animal: baby elephant] tucked under a patchwork quilt, a paper moon lamp glowing by the window, gentle snow falling outside.
STYLE: rounded soft-shape 3D animation, plush felt-like textures, pastel palette of dusty blue, cream and peach, no sharp edges anywhere.
MOTION: very slow — the quilt rises and falls with breathing, the lamp sways slightly, snowflakes drift past the window, one blink.
CAMERA: slow gentle push-in, no cuts.
AUDIO: soft music-box lullaby melody, faint wind outside, no voice.
MOOD: calm, safe, drowsy.

8 seconds, no text, nothing startling, no fast motion.

Cinematic food video of a sandwich being built.

ACTION: layers land one by one in rhythm — toasted sourdough, butter spread, folded ham, melting cheese, crisp lettuce, tomato slices, top slice pressed down, then a clean knife cut with the cheese pulling apart.
CAMERA: 50mm macro at a low 20-degree angle just above table height, slow arc around the plate, final beat pushes in on the cut cross-section.
LIGHTING: hard key from behind creating bright rim highlights on every edge, soft bounce fill in front, dark moody background.
MOTION: slow motion at the impact of each layer, real-time on the cut.
AUDIO: crunch, sizzle, knife through crust, no music.
FINISH: photorealistic, glistening textures, steam visible, ultra-sharp detail.

8 seconds, no hands visible above the wrist, no text.

Cinematic post-apocalyptic street scene.

SCENE: an abandoned four-lane city street choked with rusted cars and drifting ash, tower blocks broken against a bruised sky, one figure in a gas mask walking down the centre line away from camera.
MOTION: slow handheld follow behind the figure, ash swirling in the air, a traffic light swinging on its cable, distant silhouettes shambling between the cars in the far background.
CAMERA: 35mm, f/2.8, slight handheld drift, subject centred, deep atmospheric haze.
LIGHTING: overcast diffuse light, low contrast, one distant fire glow off frame right.
AUDIO: wind through empty streets, metal creaking, one distant guttural cry, low sub-bass drone. No music, no dialogue.
GRADE: desaturated teal and ash grey, crushed blacks, heavy grain.

8 seconds, photorealistic, single continuous take, no text, no gore.

A short dramatic film scene, photorealistic.

CHARACTERS: [A: 40s man in a rain-soaked jacket] and [B: 30s woman holding a car key].
LOCATION: a supermarket car park at night, rain falling through the sodium lamps.
BEAT: she is leaving; he arrives one second too late.
CAMERA: medium two-shot on a 50mm, slow push-in, shallow focus favouring whoever speaks, one subtle rack focus on the final line.
LIGHTING: orange sodium key from above, cold blue fill from a storefront off frame left, rain catching the light.
AUDIO: rain on metal, distant traffic, no music.
DIALOGUE:
A: "[one line, under 12 words]"
B: "[one line, under 12 words]"
Delivery: restrained, quiet, both avoiding eye contact until the last word.

8 seconds, natural lip sync, no subtitles, no text overlays.

A short dramatic film scene with Tagalog dialogue, photorealistic.

CHARACTERS: [A: nanay in her 50s, apron on] and [B: her son, early 20s, backpack on one shoulder].
LOCATION: the doorway of a small house in a Manila neighbourhood, late afternoon light, laundry lines above the alley.
BEAT: he is leaving to work abroad; neither wants to say goodbye first.
CAMERA: 35mm medium shot from inside the house looking out, slow push toward the doorway, natural handheld.
LIGHTING: warm golden hour spill through the doorway, soft shadow inside.
AUDIO: neighbourhood ambience, tricycle passing, a rooster, no music.
DIALOGUE — spoken in natural conversational Tagalog:
A: "[isang linya, wala pang 12 salita]"
B: "[isang linya, wala pang 12 salita]"
Delivery: soft, held back, voices tight.

8 seconds, accurate Tagalog lip sync, no subtitles, no text on screen.

2D stickman animation explaining a concept on a plain background.

CONCEPT: [e.g. how compound interest snowballs].
STYLE: black stick figures with thick rounded strokes on a clean off-white background, one accent colour ([#8b5cf6]) used only for the key element, simple hand-drawn icons appearing as needed.
ACTION: the stickman starts small on the left, drags a coin that doubles each step, and ends dwarfed by the final stack on the right.
MOTION: snappy pose-to-pose animation with strong anticipation and overshoot, elements pop in with a quick scale bounce.
CAMERA: static wide, one quick zoom-out at the end to reveal the full scale.
AUDIO: light pops and whooshes on each beat, upbeat minimal underscore, no voice.

8 seconds, clean and readable at phone size, no text unless numbers are essential.