Talking Head ScreenTalking Head ClassicStatic LockedSlow Burn
Luma AI: Shot-by-Shot Breakdown
All 14 cuts of Luma AI's 57-second Facebook ad on the editing timeline — framing, angle, lighting, mood and the persuasion job of every shot, plus the classification intel behind the creative.
14 cuts in 57 seconds, and not one of them moves the camera. This breakdown puts every shot of the Luma AI ad on the timeline — what is in the frame, how it is framed, and the persuasion job each cut performs while it owns the screen. Follow the playhead: all 7 beats and 14 shots below are pinned to their exact timeframe.
SRC00:00.0
Luma AI · Facebook · 0:57
Master Timeline · 7 beats · 14 cuts
▶ 00:00.0
00:00.000:57.2
Now on screen
Beat 01 · OPENING
01
Talking Head Classic
IN 00:00.0 · OUT 00:01.4 · DUR 1.4s
“Luma AI, Make it with Dream Machine, Production-ready images and videos with precision, speed, and control, Tell me more about your character..., just”Full analysis
Cut list · click a shot to jump the playhead
BEAT 01
Provocation
OPENING00:00.0 – 00:06.5 · 4 shots
It opens with a bold, extreme claim: “Luma just dropped their new creative agents and they’re insane.” The phrasing “they’re insane” is a confrontational exaggeration that signals the viewer is about to see something unusually impressive.
Why this works:This leverages Provocation by creating an immediate emotional jolt—your brain flags “insane” as a high-salience claim worth verifying. That triggers a Curiosity Spike: you keep watching to resolve whether the hype is real, because the statement sets an expectation that the next details will justify the intensity.
Visual Direction
The distinct split-screen frames the excited presenter bottom-center, actively speaking, with the Luma AI interface dynamically shown above. A dark, moody lighting strategy focuses attention on both the speaker and the glowing screen. Consistent framing establishes the presenter as the central voice, while the changing screen content visually contextualizes the product, from mission to generation options.
01
Talking Head Classic
IN 00:00.0 · OUT 00:01.4 · DUR 1.4s
“Luma AI, Make it with Dream Machine, Production-ready images and videos with precision, speed, and control, Tell me more about your character..., just”
00:00.0▶ 00:00.0 – 00:01.400:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, a digital canvas displays the Luma AI website with a 'Dream Machine' interface, showing a prompt field.
Why it works: Introduces the presenter and the AI tool immediately. The split screen effectively shows the speaker's reaction and the product he's discussing simultaneously, creating an engaging opener.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Excited And Fast-paced
Product
Luma AI website UI displayed as a large picture-in-picture screen above the speaker.
02
Talking Head Classic
IN 00:01.4 · OUT 00:02.1 · DUR 0.8s
“Luma AI, Our Mission is to build multimodal general intelligence that can generate, understand, and operate in the physical world. Product & Platf, Foundations for a new er, Creative”
00:00.0▶ 00:01.4 – 00:02.100:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website displays its mission statement.
Why it works: This shot immediately communicates the ambitious vision of Luma AI by displaying its mission statement, aiming to impress the viewer with the scope of the technology.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Excited And Fast-paced
Product
Luma AI website UI displaying its mission statement on a large screen above the speaker.
03
Talking Head Classic
IN 00:02.1 · OUT 00:02.8 · DUR 0.6s
“Luma AI, Untitled, Pre-Release & Confidential, Share, New Chat, No messages, What do you want to do?, Create, insane.”
00:00.0▶ 00:02.1 – 00:02.800:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website shows a blank canvas and a chat interface.
Why it works: Shows the blank canvas and chat interface, setting the stage for demonstrating the AI's creative capabilities and hinting at the ease of interaction with the tool.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Excited And Fast-paced
Product
Luma AI website UI displaying its main creative interface on a large screen above the speaker.
04
Talking Head Classic
IN 00:02.8 · OUT 00:06.9 · DUR 4.1s
“Luma AI, Untitled, Pre-Release & Confidential, Share, New Chat, No messages, What do you want to do?, Create, Generate Video, Generate Audio, Shapes, creates with, images,, video,, audio,, text,, all on one, canvas.”
00:00.0▶ 00:02.8 – 00:06.900:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website displays its canvas and chat, with the cursor clicking on icons for 'Generate Video', 'Generate Audio', and 'Shapes'.
Why it works: Visually demonstrates the multimodal capabilities of Luma AI by showing the user interface's options for generating various media types, emphasizing its all-in-one nature.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Instructional And Patient
Product
Luma AI website UI showing various generation options (video, audio, shapes) on a large screen above the speaker.
BEAT 02
Topic Definition
CONTEXT00:06.5 – 00:18.0 · 4 shots
It defines the core theme by stating what Luma “creates with you” and where it happens: “Luma creates with you images, video, audio, text, all on one canvas.” This immediately tells the viewer the video is about a unified, all-in-one creative workspace rather than separate tools.
Why this works:This leverages Topic Definition to reduce ambiguity early—viewers instantly know the subject and scope (“images, video, audio, text”) and the key differentiator (“all on one canvas”). By making the intent concrete, it increases cognitive fluency, so the viewer can quickly categorize the content as relevant and keep watching to see how the “one canvas” workflow works.
Visual Direction
A series of static 'END CARD' displays utilize a dark, minimalist aesthetic with clean sans-serif typography. The speaker remains present at the bottom, providing a consistent human element. Each card highlights specific Luma AI UI capabilities—from dimming to generated product variations—using clear, functional layouts, reinforced by text overlays.
05
Talking Head Classic
IN 00:06.9 · OUT 00:09.7 · DUR 2.8s
“canvas., But that's, not even the craziest, part.”
00:00.0▶ 00:06.9 – 00:09.700:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. The digital canvas above his head dims to a dark gray, then fully disappears, leaving only the speaker.
Why it works: Creates a dramatic pause and builds anticipation by momentarily fading out the product display, drawing full attention to the speaker's emphasis on the 'craziest part' of the AI's capabilities.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Confident And Engaging
06
Talking Head Classic
IN 00:09.7 · OUT 00:15.1 · DUR 5.4s
“Pre-Release & Confidential, Share, Generate Product Variations, Sneaker Product Variations, Waiting for initial sneaker and coffee variations to compl..., Launching all 6 sneaker product variation images in parallel — Studio Minimal, Neon Night, Nature..., Hide process, Coffee Product Variations, Generating..., These are, creative agents, built in, that actually, understand your entire, project, from start to, finish.”
00:00.0▶ 00:09.7 – 00:15.100:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website shows a grid of product images (sneaker variations, forest scenes, coffee beans, espresso cups) being generated.
Why it works: Illustrates the AI's ability to generate product variations and understand project context by showing multiple product images being created, highlighting its sophisticated 'creative agent' functionality.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Confident And Engaging
Product
Luma AI website UI showing a grid of generated product variations for sneakers and coffee on a large screen above the speaker.
07
Talking Head Classic
IN 00:15.1 · OUT 00:16.9 · DUR 1.8s
“Pre-Release & Confidential, Share, Generate Product Variations, Sneaker Product Variations, Waiting for initial sneaker and coffee variations to compl..., Hide process, Coffee Product Variations, Generating all 6 coffee product images in parallel with di..., Waiting for initial sneaker and coffee variations to complete before applying minimalist style direction, Minimalist style, These, agents, understand, things”
00:00.0▶ 00:15.1 – 00:16.900:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website shows the generated product variations being updated with a 'Minimalist style' text overlay.
Why it works: Demonstrates the AI's ability to apply stylistic changes to generated content, showcasing its understanding of aesthetic directions and its iterative creative process.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Instructional And Patient
Product
Luma AI website UI showing product images being processed with a 'Minimalist style' on a large screen above the speaker.
08
Talking Head Classic
IN 00:16.9 · OUT 00:20.9 · DUR 4.1s
“Pre-Release & Confidential, Share, Generate Product Variations, Coffee Ad Video Clips, Generating all 5 video clips to generate and stitch for both ads., Voiceover, music, and keyframe animation, Pickleball Sneaker Ad Video Clips, Generating all 5 video clips in parallel using keyframe animation for both ads., Final audio/voiceover, music, and text overlays, Both video clip pipelines are generating -- 10 clips total across the two ads. This is the heaviest step, then we'll do the final assembly with audio, voiceover, and text overlays., how things, move, how they, behave, in physical space,, and how light and, motion, actually, work.”
00:00.0▶ 00:16.9 – 00:20.900:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website displays sequences of generated video clips for coffee and pickleball sneakers.
Why it works: Highlights the AI's advanced understanding of physics and motion by showcasing generated video clips of products, emphasizing its ability to create dynamic and realistic content.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Instructional And Patient
Product
Luma AI website UI showing generated video clips for coffee and pickleball sneakers on a large screen above the speaker.
BEAT 03
Feature Cascade
DELIVERY00:18.0 – 00:33.0 · 2 shots
It rapidly stacks what the agents “understand” as a dense feature list: “your entire project, including how things move, behave in physical space, and how light and motion work.” Then it seals the value with a direct promise: “no need to prompt once and hope for the best.”
Why this works:This leverages Feature Cascade by making the capability feel comprehensive through rapid, specific coverage (“move,” “physical space,” “light and motion”). It also uses Assumption Shift by replacing the viewer’s default belief that you must “prompt” and “hope for the best” with the new assumption that the system handles it end-to-end automatically.
Visual Direction
Two extended 'END CARD' displays maintain the dark, minimalist visual aesthetic. Detailed UI chat logs and generated output grids are showcased through clear sans-serif typography on a moody background. The continued presence of the speaker subtly reinforces his authority as he narrates the sophisticated 'creative agent' functionalities, visually proving the AI's capabilities.
09
Talking Head Classic
IN 00:20.9 · OUT 00:27.8 · DUR 6.8s
“Pre-Release & Confidential, Share, Generate Product Variations, Let me check what's on your canvas and pull the right strategy for this., I don't see any product image on your canvas yet. Could you share a few details so I can get this right:, 1. What's the product? (e.g., a sneaker, perfume bottle, coffee bag, etc.), 2. What kind of variations? -- color/material swaps, different environments, lighting moods, camera angles, product shots, etc. Or is it a mix of everything?, 3. Any style preferences? (e.g., studio photography, lifestyle, minimalist, editorial), Or if you have a product image, drop it on the canvas and I'll build all 12 variations from that., Analyzing your request, I'll split this into two parallel jobs of 6 images each and start them., I'll create, Sneaker and coffee, Minimalist style, prompt it, once, and hope for the best., You have back and forth, conversations,, tell it what you, like,, what you don't like,, and it, iterates, with you.”
00:00.0▶ 00:20.9 – 00:27.800:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website shows a chat conversation where the user provides inputs like 'Sneaker and coffee' and 'Minimalist style', and the AI responds by planning its actions.
Why it works: Highlights the conversational nature of the AI, demonstrating how users can refine their requests through dialogue rather than single prompts, fostering a collaborative creative process.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Instructional And Patient
Product
Luma AI website UI showing the interactive chat process on a large screen above the speaker.
10
Talking Head Classic
IN 00:27.8 · OUT 00:37.1 · DUR 9.4s
“Pre-Release & Confidential, Share, Generate Product Variations, Sneaker Product Variations, Coffee Product Variations, Generating..., I asked it, to generate 12 different product variations, and it wrote, every single prompt, on its, own., It picked the right models and, kept style consistency, across, every shot.”
00:00.0▶ 00:27.8 – 00:37.100:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website displays a chat asking to 'Generate 12 different product variations', then showing a grid of generated sneaker and coffee product images.
Why it works: Emphasizes the AI's autonomy and intelligence by showing it generating diverse product variations and writing its own prompts, demonstrating its ability to maintain style consistency across multiple outputs.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Proud And Showing Off
Product
Luma AI website UI showing a chat prompt, then a grid of 12 generated product variations for sneakers and coffee on a large screen above the speaker.
BEAT 04
Feature Cascade
DELIVERY00:33.0 – 00:43.5 · 2 shots
It stacks multiple product capabilities in one breath: “back-and-forth conversations,” “it iterates with you,” “generates variations,” “picking the right models,” and “keeping style consistency across shots.” This creates a dense “everything is handled for you” value run-up, so the viewer mentally ticks off feature after feature without needing to wait for a separate explanation for each one.
Why this works:This leverages Feature Cascade to increase perceived value density—each phrase adds another concrete capability, making the system feel more complete than a single claim. It also uses Cognitive Fluency: the rapid, sequential list (“back-and-forth… iterates… generates… picking… keeping…”) is easy to process, so the viewer’s brain accepts the overall competence signal faster and stays engaged to see the full picture.
Visual Direction
Two lengthy 'END CARD' displays continue the dark, minimalist aesthetic with clean sans-serif typography. The first highlights team collaboration features and project thumbnails, while the second presents an expanded grid of generated product variations. The speaker remains constant, grounding the feature demonstrations and emphasizing Luma AI's integrated and comprehensive capabilities.
11
Talking Head Classic
IN 00:37.1 · OUT 00:42.1 · DUR 5.0s
“Adam's team, Shared With Me, MEMBER OF, An Adam's team, Manage Team, VARATIONS, Edited 24 seconds ago, Untitled, Edited 17 minutes ago, the canvas, is multi-player,, so your whole team can, collaborate, inside the same project, in, real time.”
00:00.0▶ 00:37.1 – 00:42.100:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. Above his head, the Luma AI website displays a project management interface with 'Adam's team' and 'Shared With Me' options, showing project thumbnails.
Why it works: Highlights the collaborative aspect of the platform by showcasing team management and real-time co-editing features, positioning the tool as suitable for professional team environments.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Instructional And Patient
Product
Luma AI website UI showing team collaboration features and project thumbnails on a large screen above the speaker.
12
Talking Head Classic
IN 00:42.1 · OUT 00:51.5 · DUR 9.4s
“No more, one app for, images,, another for, video,, another for, editing,, then a chat, to bring it all, together., You just, direct, the agents, orchestrate the rest.”
00:00.0▶ 00:42.1 – 00:51.500:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. The digital canvas above his head is absent, leaving only the speaker against a dark background.
Why it works: This shot emphasizes the consolidated nature of the Luma AI platform by contrasting it with the fragmented workflows of traditional creative tools, simplifying the creative process for the user.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Confident And Engaging
BEAT 05
Metric Proof
VALIDATION00:43.5 – 00:49.5 · 0 shots
It uses measurable output to validate the tool: “generate 12 different product variations” and then claims completeness with “it wrote every single prompt on its own.” This turns the viewer’s uncertainty into a concrete performance check—12 variations, 100% completion—right in the middle of the video.
Why this works:This leverages Metric Proof by anchoring credibility to specific, countable results (12 variations). It also uses a completeness signal (“every single prompt”) to trigger Certainty/Completion Bias—when the claim is absolute, the viewer feels less need to doubt and is more likely to keep watching to see if the pattern holds.
Visual Direction
BEAT 06
Overwhelm → Control
SHIFT00:49.5 – 00:54.8 · 1 shots
It reframes the usual “juggling separate apps” workflow into a controlled, unified system: “The canvas is multiplayer… collaborate in real time—no more juggling separate apps; you just direct and the agents orchestrate the rest.” This shifts the viewer from chaos-management to command-and-control, making the process feel simpler and more manageable in the moment.
Why this works:This leverages Overwhelm → Control by directly naming the pain (“juggling separate apps”) and replacing it with a single coordinated environment (“multiplayer… collaborate in real time”). The “you just direct” line reduces cognitive load, so the viewer’s brain stops tracking multiple tools and instead focuses on one clear action, which makes the solution feel immediately usable and hard to ignore.
Visual Direction
This final 'END CARD' transitions to focus solely on the speaker against a dark, unadorned background. All product UI is removed, with the message delivered via platform-native captions. This minimalist shift emphasizes the directness of the concluding statement and call to action, creating a clean, focused visual closure to the ad.
13
Talking Head Classic
IN 00:51.5 · OUT 00:57.0 · DUR 5.5s
“Creatives, have been waiting for, something like this, for a long time., If you want to check it out,, comment AI, and I'll send you the link.”
00:00.0▶ 00:51.5 – 00:57.000:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. The digital canvas above his head is absent, leaving only the speaker against a dark background.
Why it works: Creates a direct call to action, engaging the audience by inviting them to comment for more information, while reinforcing the perceived demand for such a tool among creatives.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Instructional And Patient
BEAT 07
Comment Prompt
CLOSE00:54.8 – 00:57.2 · 1 shots
It ends with a direct comment instruction: “comment AI and I’ll send you the link.” This converts passive watching into an immediate, low-effort interaction that also gates access to the resource.
Why this works:This leverages Commitment Bias and Actionability: once viewers hear “comment AI,” they can take a concrete next step without thinking, which makes the request feel easy to comply with. It also uses Social Proof—“creatives have been waiting for something like this”—to frame commenting as the obvious move for people like them, increasing the odds they’ll follow through.
Visual Direction
This final 'END CARD' maintains the focus on the speaker centered against a dark, minimalist background. Product UI remains absent, with the call to action delivered through platform-native captions. This stark, direct visual approach aims to create a personal connection, emphasizing authenticity and making the concluding instruction unambiguous.
14
Talking Head Classic
IN 00:51.5 · OUT 00:57.0 · DUR 5.5s
“Creatives, have been waiting for, something like this, for a long time., If you want to check it out,, comment AI, and I'll send you the link.”
00:00.0▶ 00:51.5 – 00:57.000:57.2
Man in a black hoodie and cap, wearing clear-framed glasses, speaking into a microphone. The digital canvas above his head is absent, leaving only the speaker against a dark background.
Why it works: Creates a direct call to action, engaging the audience by inviting them to comment for more information, while reinforcing the perceived demand for such a tool among creatives.
Angle
Eye Level Straight
Camera
Static Locked
Lighting
Warm Golden Hour
Color
Dark And Moody
Mood
Instructional And Patient
Ad Intel
The intelligence layer behind the shots — how this creative classifies, the role it plays in Luma AI’s media mix, and the structural signature that makes it identifiable at a glance.
Classification DNA
Vertical — SaaS & Software93%
Creative Format — Talking Head Screen95%
Marketing Angle — How To Tutorial62%
Intelligence Profile
Video Type
Product Demo
Platform Mode
FB Product First
Behavioral Role
Sell Product Direct
Offer State
Offer Present
0:57Duration
7Beats
14Cuts
2Cuts / Beat
2.9/10Energy
What the intel says:A Product Demo creative running in FB Product First mode with the offer present puts this at the consideration stage — it is not introducing the brand, it is closing the gap between “I've seen this” and “I'd try this.” The 2.0 cuts-per-beat rate is the tell: fast enough to stay alive on feed, slow enough to stay believable.
Ready to put this into practice?
Create your first AI UGC video ad in minutes — no filming, no actors, no editing.
