Skip to content
refs
REF-3418Source onlineBlueprint rev 01

Genie 3 World-Model Launch Film

A square launch film that proves an interactive world model through three visible tests before a rapid world-and-use-case montage.

Maker @GoogleDeepMindFor Google DeepMindOriginal post ↗
Cut timeline · 25 shots02:23
00:0000:3001:0001:3002:00

Field notes / why it works

A 13.9-second drifting-window opening poses a visual mystery before the title reveal, while the narration reframes what the viewer sees by 0:05. The film then proves live response, persistence and prompted events in separate 0:32–1:27 demonstrations; 10 cuts in roughly 19 seconds at 1:27–1:46 turn those rules into a felt range of possible worlds.

01

Plate 01 / The format in one breath

A square, narrated product launch film reframes familiar-looking moving images as something the viewer can act inside. It starts with visual intrigue, proves three capabilities through simple examples, then accelerates into a variety montage and a sober future-use ending.

Why it works

  • The first 13.9 seconds hold one drifting field of tiny worlds, then reveal a large product title over a landscape. The delayed explanation creates curiosity while the narrator names the category at 0:05.
  • Each abstract claim gets a legible proof: live response at 0:37, persistence after looking away at 0:52–1:11, and a newly prompted event at 1:12–1:27.
  • The edit spends time on proof, then speeds up sharply at 1:27–1:46: 10 cuts in roughly 19 seconds across distinct environments. This makes the product's range feel broader than a single demo.
02

Plate 02 / Format card

Reference
Platform · aspect · lengthX · 1:1, 720×720 · 2:22.5 at 23.98 fps
Pace25 measured shots; 24 measured cuts, 10.1–12.6 cuts/min including uncertain changes; 17.7 total picture changes/min; mean shot 5.7 s, median 3.21 s
Script291 spoken words; 132 wpm over full runtime, about 195 wpm during active speech; 22 sentences, 13.2 words each; narration begins at 0:01.6
VoiceOne low-register, lively, assured adult narrator; no on-camera speaker
On screenApprox. 72% generated-world / simulation scenes, 18% interface and graphic title sequences, 10% live-action operator and making-of inserts (visual estimate from sheets)
CaptionsNo running subtitles; large white feature headlines at 0:37, 0:49, and 1:13, plus small prompt UI labels
SoundContinuous low tonal/cinematic bed under narration (music-like, inferred); no clear repeated cut SFX; −14.3 LUFS integrated, −1.8 dBFS true peak
Structure7 beats: intrigue → name → three capability proofs → variety → applications → logo; no explicit spoken CTA

Pro / 13 plates + recipe

The first plates are open. The full dissection is Pro.

Get the timed shots, type and sound specs, and the runnable recipe.

Open the full blueprint
  1. Plate 03 / The hook, frame by frame (first 3 seconds)Pro
  2. Plate 04 / Structure (beat sheet)Pro
  3. Plate 05 / Shot grammarPro
  4. Plate 06 / Visual style guidePro
  5. Plate 07 / Voice and deliveryPro
  6. Plate 08 / Sound designPro
  7. Plate 09 / PackagingPro
  8. Plate 10 / Script templatePro
  9. Plate 11 / Production planPro
  10. Plate 12 / PromptsPro
  11. Plate 13 / Edit recipePro
  12. Plate 14 / QA checklistPro
  13. Plate 15 / Make it yoursPro
  14. Recipe / runnable promptPro