Video to Luma Prompt: Extract Dream Machine Prompts from Any Video (2026)
Luma’s Dream Machine speaks a very specific dialect: fluid, organic motion; naturalistic light; a dreamy, slightly surreal texture that makes ordinary shots feel like memory. Generic video-to-prompt output won’t unlock it — Luma rewards prompts that describe movement and atmosphere in natural language, not rigid shot lists.
The fix: extract the seven shot elements (subject, motion, camera, composition, lighting, style, timing) from any reference clip, then translate them into Luma’s grammar. Structure travels; the words must change.
Key takeaways: - Luma’s dialect has three pillars: motion fluidity (organic, physics-plausible movement), light naturalism (single believable light source, soft falloff), and dreamy texture (memory-like, slightly surreal grading). - Extract seven elements first, then rewrite — keeping structure but swapping vocabulary is what makes a prompt render the way the reference feels, not just what it shows. - Luma prefers concise, flowing description over stacked clauses: one subject, one camera move, one light, one mood. - Kling is action-first, Veo is audio-blocked, Luma is atmosphere-first — same structure, three dialects. - Workflow: paste link → seven elements → Luma rewrite → render → diagnose one element, revise one element. Extraction 1 credit + prompt 3 credits (≤20s clips).
What does Luma Dream Machine’s prompt dialect actually look like?
Every model has a “native” prompt style — the shape of description it follows most reliably. Luma’s dialect has three signatures:
- Motion fluency over motion volume. Where Kling wants beat-by-beat action, Luma wants flow: continuous, organic movement with plausible physics — hair drifting, fabric rippling, water sloshing, camera gliding. Name how things move, not just that they move.
- Lighting naturalism. Luma renders best when light behaves: one key light, named quality (overcast diffusion, golden-hour rim, single practical lamp), soft falloff. “Cinematic lighting” alone is wasted words; “soft overcast light from camera left, gentle falloff into shadow” is Luma food.
- Dreamy texture as mood. Luma’s signature look is memory-like and slightly surreal. Prompt words like dreamlike, hazy, nostalgic, ethereal, slow-burning align with what the model does best — and this is where most extracted prompts need translation, because reference videos rarely state their mood out loud.
How do I extract a reference video into Luma’s seven elements?
- Copy the reference clip’s link (25+ platforms supported), or upload a local MP4/MOV/WebM.
- Paste it into SocialToPrompt — extraction = 1 credit, structured prompt = 3 credits (4 total for clips up to 20 seconds).
- Receive the seven-element breakdown — subject, motion, camera, composition, lighting, style, timing — analyzed frame by frame.
- Rewrite into Luma grammar using the translation table below.
- Render, diagnose one element, revise one element. If the result feels stiff, fix motion fluency first. If it feels flat, fix lighting second. Mood last.
The general method behind step 3: how to extract a prompt from video.
How do I translate the seven elements into Luma syntax?
| Element | Extracted text (generic) | Luma rewrite |
|---|---|---|
| Subject | “a woman in a white dress in a field” | “a woman in a flowing white dress standing in tall grass” |
| Motion | “dress moves, she turns” | “the dress drifts in the wind as she turns slowly, grass swaying around her” |
| Camera | “camera pans right, then pushes in” | “the camera glides right in a slow, steady arc, then eases into a gentle push-in” |
| Composition | “centered, horizon low” | “centered figure, low horizon, wide negative space above” |
| Lighting | “golden hour” | “warm golden-hour light, soft rim glow on the dress, long soft shadows” |
| Style | “cinematic, dreamy” | “dreamlike, hazy, nostalgic film texture” |
| Timing | “6 seconds, slow” | “slow, unhurried pacing across 6 seconds, no hard cuts” |
Rules of thumb: verbs get softened (glides, drifts, eases), lighting gets qualified, style gets an emotional register, and everything flows in one paragraph rather than a checklist.
Video to Luma prompt templates you can copy
1. Dreamlike nature drift
A woman in a flowing white dress stands in a field of tall grass at golden hour. The dress and grass drift slowly in a warm breeze as she turns toward the camera. The camera glides forward in a slow, steady push-in. Warm golden-hour light with soft rim glow and long hazy shadows. Dreamlike, nostalgic, ethereal film texture. Slow, unhurried pacing across 6 seconds, no hard cuts.
2. Fluid orbit portrait
Close-up of an elderly craftsman at a wooden workbench, hands shaping clay. His hands move with slow, deliberate care, clay yielding softly under his fingers. The camera orbits him in a smooth, continuous arc from profile to three-quarter front. Single warm practical lamp from the right, soft falloff into deep shadow, dust motes in the beam. Quiet, intimate, slightly dreamlike realism. Gentle continuous motion across 5 seconds.
3. Surreal dream transition
A quiet apartment living room at dawn dissolves slowly into a forest clearing, furniture melting into tree trunks, window light becoming dappled sunlight through leaves. Everything transforms continuously, no cuts, matter flowing like slow water. Soft diffused morning light throughout, mist hanging in the air. Surreal, dreamlike, hazy and weightless. Seamless morph over 6 seconds, slow easing throughout.
4. Water and fabric physics
A dancer in a long silk dress steps into a shallow reflecting pool at blue hour. Water ripples outward in slow rings as the silk hem spreads and floats on the surface. The camera rises in a slow crane move from water level to a high angle. Cool twilight ambience with a single warm lantern glow behind her, reflections trembling on the water. Ethereal, cinematic, dreamlike texture. Slow continuous motion across 6 seconds.
How does a Luma prompt differ from a Kling or Veo prompt?
Same seven elements, three dialects — this is exactly why the same prompt renders differently across Runway, Kling and Sora:
| Luma Dream Machine | Kling | Veo | |
|---|---|---|---|
| Lead with | atmosphere + motion flow | action beats | shot + audio block |
| Motion words | organic, continuous | staged, physical | precise, sound-coupled |
| Lighting | named quality, soft falloff | material-revealing | scene-motivated |
| Style | mood register | texture register | production register |
| Extra syntax | none — flowing prose | keyframes, Omni Reference | audio/dialogue lines |
Practical test: if your prompt reads like a poem about movement and light, it’s Luma-shaped. If it reads like a fight choreography, it’s Kling-shaped. If it has a sound design line at the bottom, it’s Veo-shaped. Model-by-model extraction guides: video to Kling prompt, video to Veo prompt.
One-element-at-a-time debugging for Luma
Luma renders most failures as one broken element, not all seven:
- Stiff or robotic movement → motion fluency: add continuous verbs, name physics (drift, ripple, sway).
- Flat, video-like look → lighting: name the light source quality and falloff.
- Too “stock footage” → style: add a dreamy mood register.
- Rushed feel → timing: slow the pace explicitly (“slow, unhurried, no hard cuts”).
Change one element per render. Two changes per render and you won’t know which one worked.
FAQ
Can I convert any video into a Luma Dream Machine prompt? Yes — paste the video link into SocialToPrompt and get a structured seven-element prompt, then rewrite it in Luma’s dialect (motion fluency, lighting naturalism, dreamy mood). Works with 25+ source platforms and local uploads.
What makes a prompt “Luma-style”? Flowing natural language that leads with atmosphere and continuous motion, names light quality, and ends with a mood register — not a rigid shot list or stacked clauses.
Do Luma prompts need negative prompts or camera codes? No — Luma’s dialect is positive, descriptive prose. If a shot fails, fix the weakest of the seven elements rather than adding exclusion lists.
Should the prompt be in English? English gives the most reliable compliance on Dream Machine. SocialToPrompt can also output prompts in 25 languages if you want a localized draft.
How much does it cost to extract a prompt? 1 credit for extraction + 3 for generation = 4 credits for clips up to 20 seconds. New accounts start with 3 free credits.
Reverse-engineer the videos you love: SocialToPrompt.
Share Article