How to Make a Viral AI F1 Paddock Video

What You're Trying to Make

You've probably seen them on your timeline — a 5-12 second clip that looks exactly like a live Formula 1 world feed. Someone standing inside the Ferrari garage, oversized red headphones on, head-down focused on telemetry screens, completely composed. The lower-third reads "FERRARI GUEST." Mechanics in red Ferrari uniforms move in the soft-blurred background. The camera holds. Never cuts.

Then the caption admits the person isn't real, and the comments lose it.

This guide walks through how to make one yourself. We'll cover what photo to use, exactly what to put in the prompt, which AI models to chain together, and how to post the result so it actually gets views. By the end you'll have a generation-ready workflow you can repeat.

The Fastest Way — Use the F1 Guest Template on Starrd

The F1 Guest template is live in the Starrd library. It packages every step in this guide — the telephoto prompt, the camera-distance specifications, the "FERRARI GUEST" chyron graphic, the F1 timing tower overlay, the audio direction, the identity lock, and the team-color adaptation — into a single upload.

1. Pick a clear face photo. One person, front or three-quarter view, eyes open, neutral expression, decent lighting. A regular phone selfie beats a retouched headshot.
2. Open the F1 Guest template in the Starrd app or web library.
3. Optional — add Director's Notes if you want to override the Ferrari default (e.g., "make it Mercedes," "wear a red dress," "make it a rain race").
4. Tap Generate. Seedance 2.0 produces a 12-second F1 broadcast video in 5-10 minutes. The system maintains your face identity from the source photo through every frame.

That's the fast path. If you'd rather assemble it yourself, the rest of this guide walks through the manual workflow.

Manual Workflow — Step by Step

Step 1: Pick the Right Photo

The reference photo carries identity into the entire pipeline. Good photos in get good videos out. Bad photos get warped likenesses.

Use:

• One person, well-lit
• Eyes open, mouth closed or slight neutral expression
• Front or three-quarter angle
• Decent resolution (1024px or higher)
• Clean background OR cleanly cropped to head/shoulders

Avoid:

• Group photos (model gets confused which face to lock)
• Sunglasses or face-covering accessories
• Heavy filters, heavy makeup, beauty-mode smoothing
• Extreme angles (lying down, full profile)
• Other AI-generated photos as your reference (compounds artifacts)

A standard phone selfie in good outdoor light or by a window is ideal. Headshots from professional shoots also work if they're not heavily retouched.

Step 2: Write the Image Prompt for Nano Banana Pro

Nano Banana Pro (via the Kie API or Google's Gemini interface) generates the still image that becomes the basis of the video. Here's the structured prompt to paste in along with your reference photo:

Photorealistic 16:9 widescreen broadcast still capture inside an authentic Scuderia Ferrari Formula 1 garage during race weekend.
SUBJECT: [@image1 identity reference]. A young woman standing inside the Ferrari garage as a celebrity guest. She is wearing oversized red Ferrari team radio headphones with a small boom microphone (signature visual) and an elegant fitted black race-weekend luxury outfit — fitted black knit top or turtleneck. Preserve her actual hairstyle from the reference photo. Natural preserved makeup style. Minimal jewelry. NEUTRAL FOCUSED EXPRESSION — eyes slightly down and off-camera looking at an imagined telemetry monitor, mouth closed, NOT smiling, NOT posing, NOT looking at the camera.
ENVIRONMENT: Authentic Ferrari F1 garage interior — red and black team color palette, telemetry monitors glowing, helmets and laptops visible on benches, Ferrari pit crew and engineers in red team uniforms blurred in the background. Natural Ferrari garage LED lighting only.
CAMERA: Authentic F1 broadcast camera, 120-180mm sports telephoto lens, heavy lens compression, shallow broadcast depth of field, slight handheld stabilization.
OVERLAY GRAPHICS: Slim F1 timing tower on the left side. Small "Sky Sports F1 LIVE" overlay top right. Subtle "FERRARI GUEST" chyron in clean white sans-serif on a small dark band, lower-left.
STYLE: Ultra-realistic broadcast TV capture aesthetic. Documentary realism. Natural skin texture (no smoothing, no AI beauty look). Slight digital broadcast softness. Tiny compression artifacts. Realistic TV sharpening and noise. Must feel indistinguishable from a real F1 world feed broadcast still.
IDENTITY: Preserve the exact facial structure, eye shape, nose, lips, skin texture, proportions, hairstyle, natural appearance. No face changes. No AI beauty. No reshaping.
No additional text beyond the broadcast overlay graphics. No watermarks.

Generate at 2K resolution, 16:9 aspect ratio. Expect 1-2 attempts to get the chyron text rendering cleanly — broadcast graphics text is hard for image models and sometimes comes out as gibberish on the first try.

Step 3: Animate the Still on Seedance 2.0

Once you have a clean F1 Guest still, pass it to Seedance 2.0 along with this video prompt:

An elegant guest caught by the live F1 world feed broadcast camera inside the authentic Scuderia Ferrari Formula 1 garage during race weekend. SINGLE CONTINUOUS SHOT, NO CUTS, 12 seconds total. Slow continuous telephoto push-in from a slightly-wider opening shot to a face-focused close at the end.
IDENTITY LOCK: preserve the subject's exact facial identity from the reference image. Same face, eye shape, nose, lips, skin texture, hair color and texture. NO face changes, NO reshaping, NO AI beauty look.
CAMERA: 120-180mm sports telephoto lens. Heavy lens compression. Slight handheld stabilization drift. Natural imperfect reframing. Shallow broadcast depth of field. NO cinematic focus pulls, NO dolly moves, NO crane moves.
SUBJECT BEHAVIOR:
[00:00-04] Standing in garage, head slightly down looking at telemetry tablet at chest height OR forward at wall-mounted monitor. Neutral focused expression. Tiny breathing motion. Very subtle head movement. Soft hair movement from garage airflow. Camera begins slow push-in.
[00:04-07] Same composed focus. Eyes slowly shift to a different monitor (slight eye movement only). Hand may briefly press the Ferrari headset earcup against her ear, listening to team radio. Mouth stays closed. Camera continues pushing in. Monitor reflections begin to subtly flicker across her face.
[00:07-09] A Ferrari mechanic in red team uniform briefly crosses the foreground (silhouette blur passes through frame). Subject unchanged — still composed, eyes still off-camera. Camera holds steady on her.
[00:09-12] Camera continues push-in toward her face. Tiny micro-reaction — slight head tilt OR small eyebrow shift — as if reacting to something on screen. Push-in ends with face as center focus, eyes still off-camera, neutral composed expression. NEVER smiles, NEVER acknowledges camera.
AUDIO — AMBIENT GARAGE ONLY: Distant F1 engine roar throughout (cars on track). Ferrari garage ambience: mechanic chatter indistinct, tool sounds, headset radio static crackle (no clear dialogue), equipment hum. NO commentator dialogue, NO play-by-play, NO music. Generate audio.
STYLE: Must feel completely indistinguishable from real Formula 1 world feed footage. Documentary realism, natural skin texture, imperfect candid framing. 1080p broadcast TV feel. Subtle live TV compression artifacts. Realistic TV sharpening and noise. NO AI beauty look. NO influencer aesthetic. NO music-video framing.
AVOID: smiling at camera, eye contact with camera, posing, multiple cuts, cinematic editing, beauty filter, glam, music, commentator dialogue, picture-in-picture, embedded screens, focus pulls, sitting/seated framing (subject is STANDING), Ferrari engineer/staff appearance on the subject (the subject is a GUEST not staff — no red polo, no paddock pass, no lanyard, no team uniform).
真人实拍
Generate audio.

Generate at 12 seconds, 16:9, 720p (or 1080p if your provider supports it). Seedance 2.0 pricing on most platforms is $3 per generation for 12-second clips at high quality.

Want a deeper teardown of why every line matters? We wrote the whole F1 Ferrari Guest prompt breakdown for that.

Step 4: Pick the Right AI Video Model

The image prompt is universal. The video prompt above is written for Seedance 2.0, but the same structure works on other models with minor wording changes.

Seedance 2.0 (ByteDance, via Kie/MuAPI) — Best overall

• Reference: $3 per 12-second clip at high quality
• Strengths: prompt adherence is the best on the market right now — it actually obeys "no eye contact with camera" and "neutral expression." Identity preservation through 12 seconds is rock solid.
• Weaknesses: less cinematic dynamic range than Veo 3, but that's a feature for this trend, not a bug
• Pricing varies by access provider — see our Seedance 2.0 Complete Guide

Kling 2.0 (Kuaishou) — Original viral clips were Kling

• Reference: $2-4 per generation depending on length and quality tier
• Strengths: produces softer, more naturalistic motion than Seedance. Slightly more painterly result.
• Weaknesses: less prompt adherence — sometimes drifts into cinematic stylization even when you instruct it not to

Veo 3 (Google) — Premium option

• Strengths: stunning render quality and audio generation
• Weaknesses: leans heavily cinematic by default. Requires aggressive negative prompting to suppress its dramatic instincts. Higher CPM.

Runway Gen-4 — Decent fallback

• Strengths: fast iteration, good for testing
• Weaknesses: cinematic bias is the strongest of the four. Needs the most prompt force to land the broadcast aesthetic.

Recommendation: Use Seedance 2.0 for this format. The trend is engineered around prompt adherence, which is exactly Seedance's strength.

For a deeper model comparison, see Seedance vs Kling vs Veo.

Step 5: Common Mistakes That Kill the Illusion

After watching dozens of failed attempts, the five biggest mistakes are:

1. Putting the subject in a Ferrari engineer kit. Red polo, paddock pass, lanyard, cap. Reads as "AI dressed her up as an employee" not "AI placed her in the garage as a guest." Always use the elegant fitted black race-weekend outfit. The red Ferrari headphones are the only team item.

2. Smiling at the camera. The trend is about unawareness. The moment your subject smiles or makes eye contact with the lens, the viewer's brain reclassifies the clip from "broadcast catch" to "content." Lock in neutral focused expression in the prompt.

3. Adding commentator dialogue or music. Real F1 garage broadcast moments are ambient — engine roar in the distance, indistinct mechanic chatter, equipment hum. Booth commentary is for race action, not paddock catches.

4. Cinematic lighting prompts. "Dramatic backlighting," "rim light," "hero shot lighting" — all push the model toward music-video territory. The brief is natural garage LED lighting only.

5. Wide-angle lens or camera movement. Anything that doesn't read as a locked-off telephoto broadcast camera breaks the illusion in the first second. Stick with 120-180mm and slight handheld stabilization drift.

Step 6: Customize the Team

Ferrari is the highest-converting default because its red garage is the most recognizable F1 visual on television. But the format adapts:

• Mercedes: silver and black garage, Mercedes star branding, "MERCEDES GUEST" chyron, silver headphones
• Red Bull: navy and red garage, Red Bull branding, "RED BULL GUEST" chyron, navy headphones with yellow accents
• McLaren: papaya orange garage, McLaren branding, "MCLAREN GUEST" chyron, papaya headphones
• Aston Martin: British racing green garage, Aston branding, "ASTON MARTIN GUEST" chyron, green headphones

The Starrd F1 Guest template accepts free-form Director's Notes — "make it Mercedes" handles the entire swap automatically.

Step 7: Posting Strategy

Captions on top-performing F1 paddock AI videos follow specific patterns:

• POV framing: "POV: you got Ferrari paddock access for the Monaco GP 🏎️"
• Fake recap: "still recovering from last weekend at Imola 😭"
• Deadpan: "FERRARI GUEST is now my official job title."
• Reveal-at-end: standard hook ("here's me at the Grand Prix"), then "...except it's AI. you'd never guess right?"

Per platform rules (TikTok, Instagram Reels, YouTube Shorts), label AI-generated content using the platform's built-in disclosure tool or hashtag. The framing of your caption still matters within the disclosure rules — describe it like a broadcast moment, not as an AI tech demo.

Best hashtags for May 2026 (use 3-5, not all):

#F1 #Formula1 #FerrariF1 #ScuderiaFerrari #PaddockLife #F1Paddock #AIvideo #SkySportsF1 #AIFan #F1Trend

Why The Format Is Going Viral Right Now

A few factors are converging in May 2026:

1. F1 season is mid-stride — Imola, Monaco, and the European leg are happening, which means baseline interest in F1 broadcast content is peaking
2. Nano Banana Pro and Seedance 2.0 are both at peak quality — the technical bar for indistinguishable-from-real broadcast capture was crossed in late April
3. The TikTok / Reels algorithm rewards "wait, what?" moments — the AI reveal at the end of the clip is engineered for the algorithm's engagement signals
4. F1's mainstream cultural moment continues — Drive to Survive's afterlife, the Hamilton-to-Ferrari move from 2025, the Cadillac entry — F1 has the most mainstream attention it's ever had
5. The "GUEST" framing is socially aspirational — viewers see celebrities in paddock garages constantly; the format lets anyone insert themselves into that world

Try It on Starrd

The F1 Guest template is the fastest way to produce one of these from scratch. Upload one photo, get a 12-second F1 broadcast video — all the prompt engineering above is already baked in.

Free reads:

• F1 Ferrari Guest AI: Trend Explained — line-by-line breakdown of why every prompt element matters
• KBO Fan Cam AI: Stadium Goddess Prompt Explained — the broadcast-catch format that started everything
• Seedance 2.0 Prompt Guide — general Seedance 2.0 prompt writing

Frequently Asked Questions

How do I make an AI F1 paddock video?

Pick a clear face photo, choose an F1 team (Ferrari default), write a telephoto broadcast prompt with the "FERRARI GUEST" chyron and "Sky Sports F1 LIVE" overlay, generate the still on Nano Banana Pro and animate it on Seedance 2.0. Or use the Starrd F1 Guest template to skip every step.

What AI model is best for F1 paddock videos?

Nano Banana Pro for the image, Seedance 2.0 for the video. Seedance has the best prompt adherence — which matters because the format depends on the model NOT smiling, NOT looking at camera, NOT performing.

Why does my AI F1 video look fake?

Five common tells: Ferrari engineer kit on the subject (the trend is a GUEST not staff), smiling or eye contact with camera, "cinematic" or "dramatic" in the prompt, wide-angle lens, or added music / commentator dialogue.

Does it have to be Ferrari?

Ferrari is the default — it's the most viral. The format works for Mercedes (silver), Red Bull (navy), McLaren (papaya), Aston Martin (green). The F1 Guest template accepts Director's Notes like "make it Mercedes" to handle team swaps automatically.

Can I make this without writing prompts?

Yes. The Starrd F1 Guest template handles every step.