Higgsfield Prompts: 24 Examples with Camera Presets
24 Higgsfield prompts paired with its named camera presets, for hooks, product reveals and cinematic shots. Higgsfield is a platform that hosts several video models, so the page also explains how the preset, the model and the text prompt split the work. Sora 2 was removed from its model list after OpenAI shut down the Sora API on September 24, 2026.
★★★★★ 4.9/5 · 27,000+ users
Updated September 25, 2026.
Prompt structure
The Higgsfield prompt formula
Higgsfield's camera-control guide splits the job in two: parameters lock the technical choices (move-set, lens, focal length, aperture, lighting) and the text prompt supplies what those cannot, namely speed, path, start and end positions, and exact framing. It says variation between generations comes from whatever the words leave unspecified about the camera move. With a preset selected, the prompt stays short and focuses on the subject; with Cinema Studio, you number shots and tag references with @image and @video.
Strengths and limits
What Higgsfield is good at
Good at
- Named camera-motion presets (dozens on the Camera Controls page, e.g. Crash Zoom In, Dolly In, Bullet Time, FPV Drone, 360 Orbit, Snorricam) that you pick instead of describing the move from scratch
- Choosing between several frontier video models (Seedance, Kling, Veo, Wan) in one workspace for the same shot
- Cinema Studio: a production layer with explicit controls for genre, lighting, lens, focal length, aperture and camera move-set, plus Cast and Locations for multi-shot sequences up to 30 seconds
- Character consistency with Soul ID, trained from real photos and reusable across Higgsfield tools
- Image-to-video with Higgsfield DoP, which animates a still with a camera move while keeping face, lighting and geometry
Limits
- Limits depend on the model you pick: e.g. Seedance 2.5 up to 30 seconds at 480p-1080p, Kling 3.0 up to 15 seconds, Veo 3.1 4-8 seconds
- Some models and resolutions are gated by plan (Seedance 2.5 at 1080p is early access on Plus and above)
- Presets fix the camera move; the prompt still has to supply speed, start and end framing, and what the subject does, or results vary between generations
- Model availability changes often (Sora 2 disappeared from the picker in September 2026), so prompts should not depend on one model's quirks
- Each hosted model keeps its own content policy on top of Higgsfield's
Copy and paste
24 Higgsfield prompts to copy
UGC ad
1. Selfie review, handheld
Model: Kling 3.0 · Camera preset: Handheld A woman in her twenties with a messy ponytail and a lilac hoodie films herself in a parked car, holding a small lilac lip tint up to the lens. She swipes it on, presses her lips together and smiles. Slight natural handheld sway at phone distance, framing stays a medium close-up the whole time, no zoom. She says: "Ten hours, one coat. I'm obsessed."
Why it works: The Handheld preset gives the phone feel, and the text only adds action, framing and a line.
UGC ad
2. Unboxing with a crash zoom
Model: Seedance 2.5 · Camera preset: Crash Zoom In Top-down view of hands opening a matte black box on a white desk. Tissue paper is pulled back to reveal a pair of silver earbuds. At the moment the case is lifted, the camera crash-zooms fast from the full box to a tight shot of the earbuds, then holds still. No other camera movement. Sound: paper rustling, a magnetic click.
Why it works: The preset handles the move while the prompt times it to the reveal, as Higgsfield's guide advises.
UGC ad
3. Kitchen demo, Soul ID presenter
Model: Kling 3.0 · Character: Soul ID presenter · Camera preset: Static The presenter, a man in his thirties in a green apron, stands at a home kitchen counter and chops an onion with a black ceramic knife in fast, clean strokes, then lifts a paper-thin slice to the camera. Locked-off medium shot at chest height, bright daylight from a window on the left. He says: "Sharp enough that I don't cry anymore."
Why it works: Soul ID keeps the same presenter across ad variants, and a static frame puts all the motion in the demo.
UGC ad
4. Morning routine one-take
Model: Seedance 2.5 · Camera preset: Head Tracking A young woman in pyjamas walks from her bed to a sunny kitchen, fills a glass with water, drops in an effervescent vitamin tablet and drinks. The camera keeps her head in the upper third of frame, moving smoothly at walking pace, never overtaking her. Warm morning light, cream and orange tones. She says between sips: "This is the only thing I do before coffee."
Why it works: Head Tracking keeps the subject framed through a longer one-take, which Seedance 2.5 supports up to 30 seconds.
UGC ad
5. Street style snap
Model: Kling 3.0 · Camera preset: Whip Pan A man in a camel overcoat and white sneakers waits at a crosswalk in a busy city. The light changes; the camera whip-pans left to right following him as he steps off the kerb, landing on a full-body shot of him mid-stride. Fast pan once, then steady. Sound: traffic, a crossing signal beep.
Why it works: A single whip pan with a defined landing frame makes a punchy fashion clip without random extra moves.
product
6. Bottle orbit on wet stone
Model: Veo 3.1 · Camera preset: 360 Orbit A cobalt-blue glass water bottle stands on a wet black stone in a dark studio, water droplets beading on its surface. The camera orbits a full slow circle at bottle height, constant speed, keeping the bottle centered and the same size in frame. Single hard backlight creating a bright blue rim. No zoom. Sound: a single water drip, soft ambience.
Why it works: Constant speed, fixed framing and 'no zoom' remove the unspecified details that cause variation on an orbit.
product
7. Sneaker bullet time
Model: Kling 3.0 · Camera preset: Bullet Time A white basketball sneaker is frozen mid-air above an outdoor court at sunset, small pieces of court gravel suspended around it. The camera sweeps in a half-circle around the shoe while everything stays frozen. Warm orange backlight, long shadows, deep blue sky. Sound: a low whoosh as the camera moves, then silence.
Why it works: Bullet Time is a preset that is hard to describe in words, and the prompt only supplies the frozen moment.
product
8. Lipstick macro dolly
Model: Seedance 2.5 · Camera preset: Super Dolly In A red lipstick in a gold case stands uncapped on a pink velvet surface. The camera dollies in slowly from a medium shot to an extreme close-up of the bullet's texture, easing to a stop. Soft top light, velvet texture visible, rich reds. No zoom, no rotation. Keep it subtitle-free, no watermark.
Why it works: Super Dolly In paired with 'no zoom' avoids the dolly-versus-zoom confusion Higgsfield warns about.
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. Coffee through the object
Model: Kling 3.0 · Camera preset: Through Object In A small café table by a window. The camera starts behind a hanging plant and moves forward through the leaves to a medium close-up of a flat white in a speckled ceramic cup, latte-art heart on top. Steady speed, ending centered on the cup. Warm morning light, soft steam. Sound: café chatter, a spoon on ceramic.
Why it works: The preset handles the pass-through, and the prompt defines start object, end frame and speed.
product
10. Still to motion: skincare flat lay
Model: Higgsfield DoP (image-to-video from your product photo) · Camera preset: Dolly Out Keep the product, label and lighting exactly as in the photo. The camera slowly pulls back from a close-up of the white serum bottle to reveal the full flat lay: eucalyptus leaves, a linen towel, a small bowl of water. A leaf shifts slightly; water ripples once. Smooth, slow, no zoom.
Why it works: DoP animates the still without redrawing the product, so the label stays exactly as shot.
social hook
11. FPV into the store
Model: Kling 3.0 · Camera preset: FPV Drone A fast FPV drone flight starting high above a sunny street, diving down towards a small plant shop with a green awning, flying through the open door and between shelves of potted plants, ending on a woman at the counter who looks up and waves. One continuous flight, no cuts. Sound: drone buzz, a door bell.
Why it works: An FPV dive into a business is a proven opener for local brands, and 'no cuts' keeps it one take.
social hook
12. Dolly zoom realisation
Model: Veo 3.1 · Camera preset: Dolly Zoom In A young man at his desk looks at his laptop screen and freezes as he reads something. The background stretches behind him while his face stays the same size, over about two seconds. Harsh blue screen light on his face, dark room. He whispers: "I've been paying for this every month since 2022?"
Why it works: The Vertigo effect instantly signals shock, and the line sets up a subscription-audit product.
social hook
13. Snorricam panic
Model: Seedance 2.5 · Camera preset: Snorricam A woman in a work blazer runs down an office corridor holding a laptop, the camera fixed to her body so her face stays locked centre frame while the corridor sways behind her. Fluorescent lights, a bit of motion blur. She says, out of breath: "The deck is due in five minutes and I haven't opened it." Keep it subtitle-free.
Why it works: Snorricam's body-locked frame is attention-grabbing and sets up a productivity or AI-tool pitch.
social hook
14. Timelapse glow-up
Model: Kling 3.0 · Camera preset: Timelapse Human A small, cluttered bedroom transforms in timelapse: a person clears the floor, paints the walls sage green, builds a shelf and hangs plants. The camera stays locked in one corner as the day passes outside the window. Ends on the finished, sunlit room. Sound: fast-forward rustles, ending in a calm chime.
Why it works: A locked timelapse makes the transformation readable, a format that performs well on Reels and TikTok.
talking head
15. Founder at the window
Model: Kling 3.0 · Character: Soul ID founder · Camera preset: Static The founder, a woman in her forties with a grey blazer and short dark hair, stands by a large office window and speaks to camera. Medium close-up, eye level, locked-off. Soft daylight on her face, city softly out of focus behind. She says: "We built this because we were tired of waiting three weeks for a video."
Why it works: A static preset and Soul ID are the most stable setup for a talking head; nothing competes with the face.
talking head
16. Coach walk-and-talk
Model: Veo 3.1 · Camera preset: Dolly Out A running coach in a black windbreaker walks towards the camera along a park path at dawn while talking. The camera dollies back at her walking pace, keeping her in a medium shot the whole time. Golden low sun behind her, mist over the grass. No zoom. She says: "You don't need motivation. You need a time and a pair of shoes by the door."
Why it works: Matching the dolly speed to her walk in text keeps the framing steady for an 8-second Veo clip.
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. Doctor explainer, gentle push
Model: Seedance 2.5 · Camera preset: Dolly In A doctor in his fifties in a white coat sits at a desk in a bright clinic office and speaks to camera. The camera dollies in very slowly from a medium shot to a medium close-up over the whole clip, then stops. Soft window light, neutral tones. He says: "Eight hours isn't the goal. Waking up at the same time every day is." Keep it subtitle-free, no watermark.
Why it works: A slow push over the full clip adds intimacy without distracting from the line.
talking head
18. Podcast arc
Model: Kling 3.0 · Camera preset: Arc Right Two hosts at a podcast table with microphones in a studio with warm lamps. The camera arcs slowly to the right from a two-shot to a medium close-up of the host in the denim shirt as he speaks. Host in denim (laughing): "The first ad we ran had a typo in the price. It was our best one." Host in black (dry): "We don't talk about that." Sound: room tone, soft laughter.
Why it works: The arc moves attention to the speaker, and labelled lines keep voices on the right people.
cinematic
19. Car chase at night
Model: Kling 3.0 · Camera preset: Car Chasing Night city highway. The camera chases a black vintage sedan from behind and slightly to the left at high speed, keeping the car a steady size in frame. Neon signs streak past, wet asphalt reflects red tail lights. Sound: engine roar, tyres hissing on wet road, a distant siren.
Why it works: The preset carries a hard-to-describe vehicle move; the text only fixes distance, side and mood.
cinematic
20. Crane reveal over the harbour
Model: Veo 3.1 · Camera preset: Crane Up A fisherman in a yellow oilskin coils rope on a small wooden boat at dawn. The camera starts at his eye level and cranes up slowly and steadily to a high wide shot revealing the whole harbour, dozens of boats and gulls in the mist. Soft pink and grey light. Sound: gulls, water lapping, a distant boat horn.
Why it works: Start and end framing are both named, which is exactly what the guide says the preset cannot decide for you.
cinematic
21. Cinema Studio: rooftop duel
Cinema Studio · Genre: Action · Lighting: Contre-jour · Lens: Anamorphic · Move-set: Epic Scale Shot 1: Wide shot, two figures face each other on a rain-soaked rooftop at dusk, city behind them. Shot 2: Low-angle close-up of @image1 tightening her grip on a staff, rain dripping from her hood. Shot 3: Tracking shot as she lunges forward. Keep her black hood and scar over her left eye consistent in every shot.
Why it works: Numbered shots, an @image reference and continuity details follow Higgsfield's Cinema Studio guidance.
cinematic
22. Hyperlapse through the city
Model: Seedance 2.5 · Camera preset: Hyperlapse A hyperlapse moving forward through a Tokyo-like shopping street from dusk into night. The camera advances at a constant pace down the centre of the street while crowds blur and signs flicker on one by one. Ends facing a glowing crossing. Sound: city ambience compressed into a rising hum. Keep it subtitle-free, no watermark.
Why it works: Constant pace and a defined end point turn a preset hyperlapse into a usable opener or B-roll.
animation
23. Toy-box 3D rotation
Model: Kling 3.0 · Camera preset: 3D Rotation Stylised 3D animation, soft vinyl-toy look. A small robot mascot with round blue eyes stands on a pastel yellow pedestal and waves. The camera rotates around it in a smooth half-circle at constant speed, ending on a three-quarter view. Soft studio light, pastel background. Sound: a cheerful beep.
Why it works: A named 3D style plus the rotation preset makes a clean mascot turntable for app or brand intros.
animation
24. Paper-cut storybook
Model: Veo 3.1 · Camera preset: Pan Right Paper-cut animation, layered like a pop-up book. A little fox in a red scarf walks along a paper river under a moon made of yellow card. The camera pans slowly right with the fox at the same speed, layers moving at different depths. Warm lamp-lit paper texture. Narrator, gentle voice: "Every night, the fox walked to the edge of the map."
Why it works: A slow pan with parallax layers suits the paper style, and Veo adds the narration in the same pass.
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
- Choose the preset before writing. Then describe the subject's action and let the preset carry the camera, instead of describing a different move in text.
- Spell out speed and easing ('slow, steady, eases to a stop') and the start and end framing ('starts wide on the street, ends on her face'). Higgsfield says these unspecified details are where randomness comes from.
- Exclude competing movements explicitly: with a dolly, add 'no zoom'; with a locked POV, add 'no cuts, no zoom'.
- Don't confuse dolly (the camera moves) with zoom (the lens changes); the guide lists this as a common error.
- Test at 720p first, then re-render the keeper at full resolution. Reuse the exact same wording when you hit Recreate so you get consistent takes.
- For image-to-video that must keep the product or face exactly, use DoP with a preset; it animates the still without redrawing the subject.
- Use Soul ID or Cinema Studio's Cast for any recurring presenter, so the same face shows up in every ad variant.
- In Cinema Studio, number the shots, tag references with @image / @video, and include continuity details such as clothing, expression and sweat.
- On Seedance 2.5 inside Higgsfield, set era, genre, lighting and motion with the selectors and keep the text prompt short and focused on action.
Common mistakes
- Relying on text alone for camera movement when a preset exists, which produces a different move on every generation
- Selecting a preset and then describing a conflicting move in the prompt (Crash Zoom preset, 'slow orbit' in text)
- Vague speed words or none at all
- Writing a very long prompt on top of Cinema Studio settings that already define lighting and lens, and contradicting them
- Using a text-only character description for a recurring presenter instead of Soul ID
FAQ
Frequently asked questions
Is Higgsfield an AI model?
No. It is a platform that hosts several video models, including Seedance 2.5, Kling 3.0, Veo 3.1 and Wan 3.0, alongside its own tools such as Cinema Studio, DoP and Soul ID.
Can I still use Sora 2 on Higgsfield?
No. Sora 2 was removed from Higgsfield's model list after OpenAI shut down the Sora 2 API on September 24, 2026.
What are Higgsfield camera presets?
Named camera moves, such as Crash Zoom In, Dolly In, Bullet Time and FPV Drone, that you pick instead of describing the move in text. Not every preset works with every model.
Why does the same Higgsfield prompt give different camera moves?
Higgsfield says variation comes from what the prompt leaves unspecified. State the speed, the start and end framing, and the moves the camera must not make.
How long can a Higgsfield video be?
It depends on the model: up to 30 seconds with Seedance 2.5, up to 15 with Kling 3.0 and 4 to 8 with Veo 3.1.
Keep exploring
Related prompt libraries and pricing
Sources
- https://higgsfield.ai/blog/ai-video-camera-control
- https://higgsfield.ai/camera-controls
- https://higgsfield.ai/blog/cinema-studio-3.0
- https://higgsfield.ai/blog/seedance-prompting-guide
- https://higgsfield.ai/blog/seedance-2-5-on-higgsfield-2026
- https://higgsfield.ai/blog/Kling-3.0-is-on-Higgsfield-User-Guide-AI-Video-Generation
- https://higgsfield.ai/creator-hub/changelog
- https://higgsfield.ai/
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