Google Genie 4: What the New World Model Means for Fashion and Creative Teams

Quick answer: on 27 October 2026 Google DeepMind announced Genie 4, a "world model" that creates interactive, explorable scenes you can steer live with text: move the camera, change the weather, add people. It also makes sound (music, ambience, dialogue) and keeps a scene consistent for about two minutes, double Genie 3. Right now it's only a research preview for trusted testers, with Google AI Ultra subscribers and Gemini API partners next. For fashion and creative teams, it's a pre-visualisation tool for sets and shows, not a replacement for product-accurate photos yet.

Sources: Google DeepMind, 27 Oct 2026, Google blog, TechCrunch, The Verge.

What is new in Genie 4

• Steer in real time: type a change mid-scene ("dusk light", "camera pans left", "a model walks in") and the world responds.
• Native audio: sound is generated with the world, not added later.
• More realism: better physics and more human-like people, according to Google.
• Longer scenes: around two minutes of a consistent world.
• Gemini as director: Gemini models plan and check the scene while Genie draws it.

What it is not (yet)

• Not publicly available: there's a waitlist in Project Genie.
• Not a product photo tool: Google itself warns that details and objects can shift as scenes run longer. Logos, prints and exact garment details are not guaranteed.
• No published price or commercial terms yet.

How creatives can use world models

1. Set and venue scouting: walk through a virtual showroom, beach or runway before booking anything.
2. Show pre-visualisation: test lighting, crowd and camera path for a fashion show.
3. Mood for a campaign: share a living moodboard with a client instead of static images.
4. Pitching: show a brand what the world of the campaign feels like.

Then make the final, accurate images and clips with tools that hold the product steady.

Copy-paste brief to test when you get access

A minimalist concrete runway venue at golden hour, long glass wall, soft haze.
Camera starts wide at the end of the runway and glides forward slowly.
A model in a cream wool coat walks toward camera. Quiet ambient music, footsteps.
Keep the coat colour and silhouette identical throughout.

Then try steering: "switch to evening light", "camera moves to the side seats".

Genie 4 vs AI video tools

Video tools (Kling, Veo, Runway) give you a fixed clip you can't walk around in. Genie 4 gives you a world you explore and change live. For delivery-ready fashion content, a still-first workflow with product references is still the most reliable.

Make something today

• Build a whole show now: Runway Show Producer, pay as you go.
• Turn a still into a camera move: Cinematic Studio.
• Keep the same face and product across shots: Creative Workspace.
• Learn to write director-level briefs: Claude & ChatGPT workshop, or save with a workshop bundle.

FAQ

What is Google Genie 4?

Genie 4 is Google DeepMind's world model, announced on 27 October 2026. It generates interactive, explorable worlds with sound that you can change live with text prompts.

Can I use Genie 4 now?

Not yet for most people. It's a research preview for trusted testers. Google says it's coming soon to Google AI Ultra subscribers in Project Genie and to early Gemini API and AI Studio partners, and there's a waitlist.

Is Genie 4 good for fashion product photos?

Not as a final tool. Google notes details can shift over longer scenes, so exact garments, logos and prints aren't guaranteed. Use it for ideas and pre-visualisation.

How long can a Genie 4 scene last?

About two minutes of consistent world, double Genie 3's limit.

How is Genie 4 different from AI video generators?

Video generators output a fixed clip. Genie 4 creates a world you can explore and steer in real time, with audio generated at the same time.