VOYAGEFILMMAKERS ®
Start a project
CASE STUDY POST · 02 9 MIN READ

How we made the Bugatti Tourbillon ad with AI (in 3 weeks).

In late 2024, InVideo brought us a brief that, on paper, was impossible. Make an AI film for the Bugatti Tourbillon — a hypercar that had just been announced. No public footage. No real shots. No press images beyond a few angles.

And the kicker: feature Kendall Jenner driving it through Iceland.

We had three weeks. We shipped 33.3K views and a campaign that ran across InVideo's channels. Here's exactly how we built it.

A · THE PROBLEMWhy the brief looked impossible

AI had no idea what a Tourbillon looked like.

Every AI video model is trained on what already exists on the internet. The Bugatti Tourbillon had launched maybe a month before we started. So when we typed "Bugatti Tourbillon" into any generator, we got Chirons, Veyrons, and weird hybrids. The model had no real reference.

Akanksh and I had two choices: wait for the model to catch up (it wouldn't, in 3 weeks), or build the look from scratch. We built it from scratch.

STEP · 01The Bugatti character sheet

Lock the car before you shoot it.

A character sheet in AI work is the same thing it is in animation. It's a multi-angle reference of one subject so the model knows what to render every time you ask for it. For the Tourbillon, we built a sheet across seven specific angles.

  1. 01
    3/4 front angleThe hero angle — set the proportions, body lines, and the signature bisected headlight cluster as the anchor for every other view.
  2. 02
    Side profileLocks the silhouette — wheelbase, roofline, the long-hood / short-rear hypercar geometry.
  3. 03
    Rear 3/4 angleSets the rear-light signature and exhaust geometry — easy place for the model to drift.
  4. 04
    Top-downTop plate for any aerial / drone-equivalent shot. Hard to fake later if you skip it.
  5. 05
    Interior dashboard viewThe Tourbillon's mechanical Swiss-watch cluster is the brand moment. Had to be exact.
  6. 06
    Headlight detailMacro reference so any hero close-up reads as Tourbillon, not generic Bugatti.
  7. 07
    Wheel detailCarbon-fiber weave + spoke design lock — the second-most-frequent "tell" after the headlight.

Each angle took 15 to 25 generation passes to get right. We used Bugatti's official press shots as anchors, then iterated until the proportions, body lines, and signature details were dead consistent.

Total images burned to lock the car: about 350. Total kept for the character sheet: 8.

STEP · 02The Kendall Jenner character sheet

Same process. Harder problem.

AI models are trained on celebrity images, so they'll generate "a Kendall Jenner lookalike" — but they break down the moment you ask for specific angles, expressions, or lighting conditions.

We built her sheet around six anchor frames:

  • Driving position — hands at 10 and 2, eyes on the road.
  • Side profile in motion — for cutaways during the drive sequence.
  • Smiling at camera — low angle, hero shot.
  • Behind the wheel — interior, framed against the dashboard.
  • Walking away from the car — full body, scale reference.
  • Close-up of eyes and mouth — the brand moment.

Took us about 250 images to lock her character sheet. The hardest part: getting consistent eye colour, jaw structure, and hair behaviour across angles.

STEP · 03Iceland

The easy part.

The model knows Iceland. Black sand beaches, basalt columns, glacial lagoons, the works. We pulled together a location moodboard from real Iceland photography (Diamond Beach, Stokksnes, Vík), then generated environments to match. Four locations made it into the film:

  • Reynisfjara black sand beach — driving sequence.
  • Stokksnes with Vestrahorn mountain — hero shot.
  • Diamond Beach — Kendall walking up to the car.
  • A long basalt road through fog — the closing shot.

Iceland took us about 200 environment generations. Less than the car or Kendall because the model already had the data.

STEP · 04Putting them together

Where most AI work falls apart.

Once we had the Bugatti sheet, the Kendall sheet, and the Iceland environments, we started building the shots. Dropping a character into an environment usually breaks one or the other — so we ran a stacked pipeline:

  • Reference image conditioning — feed the model the car sheet + Iceland environment together.
  • Multi-image input — Seedance 2.0 lets you reference up to 12 files at once.
  • Img-to-vid generation for motion — Veo 3.1 and Kling 3.2 for the actual movement.
  • Manual frame fixes — Photoshop and After Effects where the AI broke.
// THE NUMBERS

What three weeks looks like.

900+ Images generated
100+ Video clips generated
40+ Models & tools used
3 wk Brief to delivery
STEP · 05The edit

The part nobody talks about.

AI doesn't edit your film. We do. Akanksh sat for five days in the edit timeline — colour grading the AI footage so it matched a unified look. Adding camera shake and grain so it felt less synthetic. Cutting on motion to hide the model breaks.

Sound design used real Bugatti engine audio from the manufacturer's reveal video, layered with synthesized low-end. Iceland ambient (wind, distant waves) underneath. A custom score on top.

Without the edit, AI footage looks like AI footage. With the edit, it looks like a Bugatti commercial.

// WHAT WE LEARNED

Three things we now apply to every car shoot.

These showed up first on Tourbillon. They show up on every car project since.

⌖ LESSON 01

Build the character sheet first, always.

Even if the brand thinks they want shots fast, building the sheet up front saves 3× the time downstream. Skip it and you'll burn the same hours later — just under more pressure.

⌖ LESSON 02

Real audio beats synthetic.

Use real engine sound, real environment sound. AI sound is the giveaway. Audiences forgive imperfect visuals before they forgive a synthetic engine note.

⌖ LESSON 03

Colour-grade like it's a real film.

Don't trust the AI's colour output. Always grade in DaVinci or Premiere afterward — it's the single biggest move that pushes a film out of the uncanny valley.

// FREQUENTLY ASKED

Five questions we get on car projects.

Q.01 How long does an AI car ad take to produce?

Two to four weeks for a 30–60 second spot, depending on whether the car has existing AI training data or needs a custom character sheet built from scratch.

Q.02 Can AI generate any car accurately?

Yes, but only if you build a character sheet. Out of the box, most AI models generate generic or wrong versions of specific cars. The character sheet is what locks the look.

Q.03 How many images does it take to lock a car character sheet?

200 to 400 image generations to land 6 to 8 final reference angles. Cars with simpler geometry need fewer. Hypercars and custom-bodied cars need more.

Q.04 Did we have permission to use Kendall Jenner's likeness?

This was a partnership project with InVideo for demonstration purposes. For commercial use of any celebrity likeness, brands must secure direct licensing rights.

Q.05 What AI tools did we use for the Bugatti project?

A combination of Midjourney, Flux, Veo, Kling, and Seedance (early versions). The tools have evolved a lot since — see our 2026 comparison post for the current stack.

// READ NEXT

Keep going.

All posts
// START A PROJECT

Want a film at this level for your brand?

Hypercars, hero products, impossible locations. Tell us what you'd shoot if logistics weren't a problem — we'll quote it within 24 hours.

Book a discovery call
WhatsApp