Sora Prompts: 24 Sora-Style Prompts for Veo 3.1 and Kling
24 Sora-style video prompts written with the structure from OpenAI's Sora 2 prompting guide: scene, cinematography, timed actions, dialogue and sound. Sora itself is shut down, so each prompt is meant to run on Veo 3.1, Kling 3.0 or Seedance, all available on Hooked. The page also covers the formula, tips and common mistakes.
★★★★★ 4.9/5 · 27,000+ users
Updated September 25, 2026.
Prompt structure
The Sora prompt formula
OpenAI's cookbook guide says to write a prompt the way you would sketch a shot on a storyboard: a prose scene description first, then a cinematography block with framing and mood, then a short list of action beats, then dialogue in its own block. Duration, resolution and character references were API parameters, so the prompt spent its words on subject, motion, light and style. Specific prompts gave consistency; lighter prompts left room for the model to vary.
Strengths and limits
What Sora is good at
Good at
- Video and audio generated together: dialogue, sound effects and ambience come out with the picture
- Following storyboard-style direction: shot framing, camera move, lighting and timed action beats
- A wide style range, from handheld phone footage to 35mm film looks to animation
- Keeping a character consistent across clips with uploaded characters (up to two per generation through the API)
- Controlled iteration: the edit endpoint changed one thing and kept the rest ("same shot, switch to 85mm")
Limits
- No longer available: the API and all Sora 2 model snapshots were shut down on September 24, 2026
- Duration was an API parameter, not something the prompt controlled: 4, 8 and 12 seconds at launch; later OpenAI docs also list 16 and 20 seconds
- Sizes: sora-2 rendered 720x1280 and 1280x720; sora-2-pro added 1024x1792, 1792x1024 and 1080p (1920x1080, 1080x1920)
- API content rules: only content suitable for under-18 audiences; no real people, including public figures; copyrighted characters and copyrighted music rejected; input images with human faces rejected
- Consumer outputs carried a visible moving watermark and C2PA metadata
- OpenAI's guide notes the model follows instructions more reliably in shorter clips; complex motion over longer durations drifted more
Copy and paste
24 Sora prompts to copy
UGC ad
1. Bathroom serum review
Vertical phone video in a small sunlit bathroom. A woman in her late twenties with freckles and a loose bun holds a frosted green serum bottle close to the lens. Cinematography: selfie framing, medium close-up, slight handheld sway, natural window light, soft greens and warm skin tones. Mood: candid, unpolished. Actions: - She tilts the bottle so the label catches the light. - She presses one drop onto the back of her hand and rubs it in. - She looks back into the lens and grins. Dialogue: Woman: "Three weeks in and I stopped wearing foundation. That's the review." Sound: bathroom fan hum, a soft click as the dropper closes.
Why it works: The selfie framing, handheld sway and short line read as real UGC, and the three beats fit an 8-second clip.
UGC ad
2. Car-seat coffee rant
Vertical phone footage from the passenger seat of a parked car on an overcast morning. A man in his thirties in a grey hoodie holds a matte black travel mug. Cinematography: phone propped on the dashboard, medium shot, slightly low angle, flat daylight, muted blues and greys. Mood: tired, then pleasantly surprised. Actions: - He takes a sip and pauses. - He turns the mug to show it is still steaming. - He points at it with his free hand. Dialogue: Man: "Poured this at six. It's nine-forty. Still hot. I'm keeping it." Sound: rain ticking on the roof, a turn signal clicking once.
Why it works: A dashboard-mounted phone angle and a concrete time claim give the proof-style hook that UGC ads rely on.
UGC ad
3. Unboxing on the bedroom floor
Top-down phone video on a cream shag rug. Two hands with chipped pink nail polish open a pale blue shipping box with a pull tab. Cinematography: overhead shot, locked-off, soft afternoon light from a window on the left, palette of cream, pale blue and gold. Mood: excited, tactile. Actions: - The tab rips cleanly and the lid lifts. - The hands pull back tissue paper to reveal gold wireless earbuds in a case. - One hand snaps the case open, then shut. Dialogue: Voice off camera: "Okay, the snap alone is worth it." Sound: paper rustle, a crisp magnetic snap.
Why it works: Overhead unboxing framing plus a satisfying sound cue is a proven short-form format and fits Sora's strength with synced foley.
UGC ad
4. Gym bag checklist
Vertical handheld video in a bright apartment hallway. A woman in running gear kneels by an open charcoal gym bag and packs it while talking to the phone. Cinematography: medium shot at eye level from a phone leaning on a shoe rack, cool daylight, charcoal, white and neon-orange accents. Mood: brisk, practical. Actions: - She drops in a folded towel and a water bottle. - She holds up a neon-orange shaker and taps it twice. - She zips the bag in one pull and stands. Dialogue: Woman: "Towel, bottle, and the one thing I never forget anymore." Sound: zipper, keys jingling on a hook.
Why it works: Packing routines are relatable, and the hero product is set apart by colour and a deliberate beat.
UGC ad
5. Kitchen hack reaction
Vertical phone video in a small, slightly messy kitchen at night. A young man in a faded band t-shirt holds a silicone lid over a bowl of leftover pasta. Cinematography: medium close-up, handheld from a friend's phone, warm overhead light, orange and olive tones, light grain. Mood: skeptical, then impressed. Actions: - He presses the lid onto the bowl and it seals with a pop. - He flips the bowl upside down and nothing spills. - He looks at the camera, eyebrows raised. Dialogue: Friend off camera: "No way." Man: "Five bucks. Five." Sound: the suction pop, a fridge humming.
Why it works: The upside-down test is a visual proof the model can stage in three beats, and the off-camera voice adds authenticity.
product
6. Perfume on marble
A faceted amber glass perfume bottle stands on white marble veined with grey, a single dried orange slice beside it. Cinematography: macro close-up, 100mm lens, very shallow depth of field, slow orbit of about 30 degrees to the right. Hard sidelight from the left through a window blind, casting striped shadows. Palette: amber, white, charcoal. Mood: quiet, expensive. Actions: - The stripes of light drift slowly across the bottle. - A drop of condensation slides down one facet. Sound: near silence, a faint room tone, one soft chime at the end.
Why it works: One slow camera move, one motion detail and a restrained palette is the kind of controlled shot the cookbook recommends.
product
7. Sneaker drop on concrete
A white and volt-green running shoe falls into frame and lands on wet black concrete at night. Cinematography: low angle, ground-level close-up, 35mm lens, shallow depth of field, locked-off camera. Sodium street light behind the shoe creates a warm rim; a cool blue fill from the left. Mood: punchy, urban. Actions: - The shoe hits the ground and water splashes outward in a ring. - Droplets hang, then fall. - The laces settle. Sound: a heavy thud, splash, distant traffic hiss.
Why it works: A single physical event with clear sound cues plays to Sora's physics and audio sync.
product
8. Skincare shelf reveal
A pastel bathroom shelf holds four skincare jars in sage, blush, sand and white. Cinematography: medium shot, slow push-in from shelf width to the centre sage jar over the whole clip, 50mm lens, soft diffused morning light, gentle bloom on highlights. Mood: calm, clean. Actions: - A hand enters from the right and turns the sage jar so its label faces the lens. - The hand leaves frame and the push-in finishes on the label. Sound: a light ceramic clink, soft birdsong through an open window.
Why it works: The push-in is tied to the clip length and ends on the label, so the product is the last thing viewers see.
Run these prompts without five subscriptions
- Paste the prompt into Veo 3.1, Kling 3.0, Seedance 2.0, Nano Banana Pro and GPT Image 2, all in one place
- Turn the clip into a TikTok, Reel or Short with captions and a hook
- Publish straight to TikTok, Instagram and YouTube
product
9. Headphones in the rain
Over-ear headphones in matte sand colour hang from a hook by an open window while rain falls outside. Cinematography: close-up, 85mm lens, rack focus from raindrops on the glass to the headphones at mid-clip. Overcast daylight, cool greys with warm sand tones. Mood: introspective. Actions: - Raindrops run down the glass. - Focus shifts to the headphones as they sway slightly on the hook. Sound: rain on glass, which cuts to muffled silence the moment the focus lands on the headphones.
Why it works: The sound drop is the product claim (noise cancelling) told through audio, which Sora generates in the same pass.
product
10. Chocolate bar snap
A dark chocolate bar in a matte terracotta wrapper lies half-unwrapped on a walnut board. Cinematography: extreme close-up, 100mm macro, locked-off, warm tungsten key from the upper right, deep shadow on the left. Palette: terracotta, walnut brown, cocoa. Mood: indulgent. Actions: - Two fingers grip the exposed end. - The bar snaps cleanly in one motion; a few crumbs scatter. - The broken edge is held up to the light. Sound: a sharp, clean snap and the rustle of foil.
Why it works: A single satisfying snap with matching sound is a strong loopable product moment.
social hook
11. Stop-scrolling spill
Vertical phone video in a sunny café. A barista slides a full latte across the counter toward camera and it tips over the edge. Cinematography: counter-level close-up, handheld, fast, bright daylight, cream and espresso tones. Mood: sudden. Actions: - The cup slides and tips in the first second. - A hand shoots in and catches it mid-air without spilling. - The hand sets it down upright. Dialogue: Barista: "Sorry. Anyway, here's why we switched to these cups." Sound: ceramic scrape, gasp from someone off camera.
Why it works: The action lands in the first second, which is what a scroll-stopping hook needs, and the line pivots straight to the pitch.
social hook
12. Wrong way, right way
Vertical split-feel video in a kitchen. A woman in an apron faces the camera holding a knife over an avocado. Cinematography: medium close-up, static tripod, bright flat daylight, green and white palette. Mood: playful, instructional. Actions: - She stabs toward the pit and stops. - She shakes her head at the camera. - She sets the knife down and twists the avocado halves apart by hand. Dialogue: Woman: "Stop doing this. Do this." Sound: a knife tapping the board, a soft squelch as the halves separate.
Why it works: A 'stop doing this' opener with a visible correction makes the hook and the payoff land in under eight seconds.
social hook
13. POV: first day at the new desk
First-person POV walking into a sunlit co-working space and sitting at an empty desk by a window. Cinematography: POV at eye level, smooth handheld, 24mm lens, bright morning light, white, oak and plant green. Mood: fresh start. Actions: - The viewer's hands set down a laptop and open it. - A coffee cup is placed to the right. - The hands pause over the keyboard. Dialogue: Voiceover, a woman's voice: "Day one of doing the thing I kept postponing." Sound: chatter, keyboard clicks, a chair rolling.
Why it works: POV framing and a relatable voiceover give a hook that fits many products and niches.
social hook
14. The 3-second test
Vertical close-up of a man's hand holding a stopwatch in front of a whiteboard marker drawing. Cinematography: close-up, locked-off, cool office fluorescent light, white and red palette. Mood: tense, curious. Actions: - The thumb clicks the stopwatch. - The other hand wipes the marker drawing with a dry cloth in one swipe; it vanishes. - The stopwatch stops at 2.9 seconds. Dialogue: Man off camera: "Permanent marker. Under three seconds." Sound: stopwatch click, squeak of cloth, final click.
Why it works: A timed challenge gives viewers a reason to watch to the end, and the stopwatch is a legible visual proof.
talking head
15. Founder at the stockroom
A founder in her forties with short silver hair stands in a small stockroom lined with kraft boxes, speaking to the camera. Cinematography: medium shot, eye level, 35mm lens, static tripod, warm practical light from a hanging bulb plus soft window fill. Palette: kraft brown, black, warm white. Mood: honest, direct. Actions: - She rests one hand on a box. - She taps the box on the key word. Dialogue: Founder: "We pack every order in this room. When something goes wrong, it's my phone that rings." Sound: faint tape gun somewhere off camera, room tone.
Why it works: A real-looking workspace and one gesture tied to a word make a founder message credible.
talking head
16. Dentist myth-buster
A friendly dentist in his fifties in navy scrubs sits on a stool in a bright clinic, speaking to camera. Cinematography: medium close-up, eye level, static, soft overhead clinic light with a warm bounce, clean whites and navy. Mood: reassuring. Actions: - He holds up a toothbrush. - He mimes brushing too hard, then shakes his head. Dialogue: Dentist: "Harder isn't cleaner. Hold it like a pencil, not a hammer." Sound: quiet clinic ambience, a distant chair motor.
Why it works: A single memorable line plus a visual demonstration fits the few-sentences dialogue guidance.
Run these prompts without five subscriptions
- Paste the prompt into Veo 3.1, Kling 3.0, Seedance 2.0, Nano Banana Pro and GPT Image 2, all in one place
- Turn the clip into a TikTok, Reel or Short with captions and a hook
- Publish straight to TikTok, Instagram and YouTube
talking head
17. Chef's one tip
A chef in a white jacket stands at a steel pass in a busy restaurant kitchen during service. Cinematography: medium shot, 50mm lens, slight handheld, warm heat-lamp glow on her face, steel and copper tones, background soft and busy. Mood: confident, quick. Actions: - She wipes the rim of a plate with a cloth. - She looks up at the lens. Dialogue: Chef: "Salt the pasta water like the sea. That's the whole secret." Sound: pans clattering, a ticket printer, someone calling 'service'.
Why it works: Background kitchen sound and a single line build context without extra dialogue.
talking head
18. Coach on the track
A track coach in his sixties in a red windbreaker stands trackside at dawn, speaking to camera. Cinematography: medium close-up, 85mm lens, shallow depth of field, low golden backlight with a soft reflector fill, red and gold tones against a blue sky. Mood: gruff, warm. Actions: - He blows into his hands against the cold. - He points past the camera toward the track. Dialogue: Coach: "Nobody remembers the fast days. They remember you showed up on this one." Sound: wind, distant spikes on the track, a whistle.
Why it works: Backlit dawn light and a short motivational line suit fitness brand content.
cinematic
19. Night market chase
Shot on 35mm film in a crowded night market in the rain. A teenage courier in a yellow raincoat weaves between food stalls carrying a paper bag. Cinematography: tracking shot beside her at shoulder height, 40mm anamorphic, neon reds and greens reflecting on wet pavement, steam from grills. Mood: urgent. Actions: - She ducks under a hanging lantern. - She sidesteps a vendor pushing a cart. - She glances back once, then runs out of frame. Sound: sizzling woks, rain, a vendor shouting, her footsteps splashing.
Why it works: One tracking move and three clear beats give motion without overloading a short clip.
cinematic
20. Lighthouse keeper
A lighthouse keeper in a heavy wool coat climbs the last steps of a spiral staircase at dusk. 1970s film look, soft grain. Cinematography: low angle from the stairwell looking up, slow tilt up following him, 28mm lens. Cold blue dusk through small windows, warm lamp in his hand. Palette: slate blue, brass, amber. Actions: - He reaches the top and sets the lamp down. - He pulls a lever; the great lens begins to turn. Sound: boots on iron steps, wind, a low mechanical groan as the lens starts rotating.
Why it works: Style is set in the first sentence and a single tilt carries the shot, which the guide recommends.
cinematic
21. Desert road stop
Wide establishing shot of an empty desert highway at noon. A dusty red pickup pulls onto the shoulder and stops. Cinematography: wide shot, eye level, locked-off, 24mm lens, harsh overhead sun, heat shimmer, palette of ochre, faded red and pale blue. Mood: still, lonely. Actions: - Dust drifts past the truck. - The driver's door opens; a boot touches the gravel. - The door stays open; nobody else moves. Sound: engine ticking as it cools, wind, a single crow.
Why it works: A static frame lets small actions and ambient audio carry tension.
cinematic
22. Last train
A woman in a green coat sits alone in a late-night subway car, forehead against the window. Shot like a 1990s Hong Kong film, step-printed motion blur. Cinematography: medium shot across the aisle, 35mm, flickering fluorescent overhead light, tunnel lights streaking past. Palette: sick green, amber, black. Actions: - The lights flicker twice. - She opens her eyes and looks at her reflection. - She smiles slightly. Sound: rails clacking, a muffled announcement, the flicker buzz.
Why it works: Naming a film-era look anchors grading, blur and pacing at once.
animation
23. Clay bakery morning
Stop-motion claymation of a tiny bakery at sunrise. A round clay baker with a flour-dusted apron pulls a tray of bread rolls from a brick oven. Cinematography: medium shot, slightly high angle, static, warm orange light from the oven, pastel blue morning light through the window. Visible fingerprints on the clay, 12 frames per second feel. Actions: - The oven door swings open with a puff of steam. - He slides the tray out; the rolls wobble. - One roll hops off and rolls away. Sound: a creaky hinge, a soft 'boing', a surprised hum.
Why it works: A defined animation medium plus a comic beat shows the style range the model supports.
animation
24. Paper city explainer
Flat 2D cut-paper animation of a small city. A paper delivery van drives across a folded street and stops at a paper door. Cinematography: side-on view like a diorama, slow pan right following the van, even soft light, paper textures visible. Palette: mint, coral, cream, navy. Actions: - The van stops; a paper parcel slides out onto the doorstep. - The door opens and a hand takes it. - A small tick icon pops above the door. Sound: paper rustling, a cheerful bell ding.
Why it works: A simple visual language and a single pan make an easy loop for app or delivery explainers.
From prompt to posted video
Paste the prompt, get the video, post it
Hooked Pro is $29 a month for 1,000 credits across Veo 3.1, Kling 3.0, Seedance 2.0, Nano Banana Pro and GPT Image 2, plus AI UGC ads, captions and TikTok slideshows in the same plan. Or pay $89 once for the Studio and run it on your own API keys, at the provider's price.
Better results
Tips that make a difference
Do
- Name the shot the way a cinematographer would: 'medium close-up, slow push-in' or 'wide establishing shot, eye level', not 'a nice shot of'.
- Break motion into counted beats: 'takes four steps to the window, pauses, pulls the curtain in the final second' instead of 'walks across the room'.
- Put dialogue in its own block under the scene, and keep it to a few short lines that fit the clip length. A 4-second clip holds one or two sentences.
- Anchor the palette with 3-5 named colours and describe the light sources ('soft window key, warm lamp fill, cool rim from the hallway') instead of 'well lit'.
- Use sound as a rhythm cue: 'distant traffic hiss', 'a crisp snap as the lid closes'. Describe diegetic sound, not a whole soundtrack.
- Set the style early in the prompt ('shot on 16mm', 'vertical phone footage', 'IMAX-scale') because it frames every choice after it.
- Prefer two short clips stitched together over one long clip; the guide says instruction-following is more reliable in shorter generations.
- When iterating, change one variable at a time ('same shot, switch to 85mm', 'same shot, golden hour') so you can tell what moved the result.
- If you use an image as the first frame, match it to the output resolution so the composition locks in.
Common mistakes
- Writing 'make it 20 seconds' or 'in 4K' in the prompt: length and size were set in API parameters, and the model ignored them in prose
- Cramming several camera moves and scene changes into one short clip, which the model resolved by dropping some of them
- Long dialogue that cannot fit the duration, which gets rushed or cut off
- Vague adjectives ('cinematic', 'beautiful') without the framing, light or lens that would produce them
- Prompting real people, public figures or copyrighted characters, which the API rejected
FAQ
Frequently asked questions
Is Sora still available?
No. OpenAI closed the Sora app on April 26, 2026 and shut down the Sora 2 API, including sora-2 and sora-2-pro, on September 24, 2026.
What replaced Sora?
OpenAI named no replacement. The closest models are Veo 3.1, Kling 3.0 and Seedance, which also generate video with native audio and follow shot-by-shot direction.
Do Sora prompts work on other models?
Yes, with small changes. The structure (scene, camera, timed action beats, short dialogue, sound cues) maps onto Veo 3.1, Kling 3.0 and Seedance. Set duration and aspect ratio in each model's settings, and check its syntax for dialogue and sound.
How long could a Sora 2 video be?
Duration was an API parameter, not prompt text. At launch it was 4, 8 or 12 seconds; later OpenAI docs also list 16 and 20 seconds.
Why did my Sora prompt get rejected?
The API refused real people, including public figures, copyrighted characters and music, and content not suitable for under-18 audiences. Input images with human faces were also rejected.
Keep exploring
Related prompt libraries and pricing
Sources
- https://help.openai.com/en/articles/20001152-what-to-know-about-the-sora-discontinuation
- https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide
- https://developers.openai.com/api/docs/guides/video-generation
- https://developers.openai.com/api/docs/deprecations
- https://openai.com/index/sora-2/
- https://openai.com/index/launching-sora-responsibly/
- https://cdn.openai.com/pdf/50d5973c-c4ff-4c2d-986f-c72b5d0ff069/sora_2_system_card.pdf
- https://en.wikipedia.org/wiki/Sora_(text-to-video_model)
Two ways in · pick yours
Every template here opens in the app — 1,600+ and counting.
Pay monthly and we run the AI for you, credits included — or pay $89 once, own the app for life and bring your own keys.
Pro
Most flexibleWe run the AI for you. No API keys, nothing to set up — open the app and generate.
1,000 credits a month · cancel anytime
Secure checkout by Stripe · Cancel anytime
- About 10 UGC ads or 250 images a month
- Every video format: UGC, avatars, slideshows…
- 83 AI avatars & 76 voices in 25 languages
- Publish to TikTok, Instagram & YouTube
- 3 channels, 3 auto-channels, API included
Need more? One-time credit packs from $19. They never expire.
4.9/5 · 27,000+ users create with Hooked