Google’s Genie 4 Builds Talking Worlds in Real Time – Waitlist Only
In short
- Google DeepMind announced Genie 4 on October 27: a world model that builds a steerable world in real time, now with music, ambient sound and dialogue.
- Access today: research preview for trusted testers. Consumers and developers come later, starting with Google AI Ultra subscribers in Project Genie and early Gemini API and AI Studio partners. There is a waitlist.
- Money angle: no price announced, so for now it is a demo to watch, not a tool to budget for.
Google built a world you can talk to. You type “make it rain” and it rains. You add a person and that person speaks. Impressive. Can you use it today? No.
What happened
On October 27, Google DeepMind announced Genie 4, its new world model. Earlier Genie versions made worlds you could move through. This one is natively multimodal: music, ambient noises and interactive dialogue are generated together with the world itself.
How does it work? Genie 4 builds on the Gemini family. Gemini Omni anchors the simulation in real-world appearance and behavior. Other Gemini models keep the scene coherent while it runs. You steer with standard keyboard controls for movement and camera, and with text prompts: change the weather, move the camera, add a person or an object. Google says it also wants these worlds for training environments for robots and agents.
Who gets it? Google calls this a research preview for trusted testers, to collect feedback before a wider release. Consumers and developers come next, starting with Google AI Ultra subscribers in Project Genie and early access partners of the Gemini API and Google AI Studio. Everyone else can join the waitlist. No price. No date.
How to use it
Today, honestly: join the waitlist and wait. If you build games, ads or simulations, that is the only step that makes sense. Here are the uses I would test once access opens (my ideas, not Google’s):
- Pitch demos. Show a client a playable scene instead of a mood board.
- Ad concept tests. Try five scenes before paying a studio for one.
- Training scenarios. Google itself points at robot and agent training environments.
Do the math: there is no price, so I’ll use the cost it could replace. Say a freelancer charges €300 for one playable concept demo, and you pitch ten clients a year. That is €3,000. If Genie access ends up costing less than that, the numbers work. Nobody knows yet, so don’t plan around it. These are my example numbers.
What it means
World models are moving from silent demo clips to full scenes with sound and speech. Google frames it as a step toward AGI. For builders the useful part is simpler: a scene you can edit with a sentence. If it ever gets a price and an API, game and ad studios gain speed first. Until then, Google gives you a blog post, videos and a waitlist.
Ivan’s Take
Cool tech, zero use today. I won’t build a plan on a waitlist. I’d sign up, wait for a price and an API, then run one test: can it make a client pitch demo in an afternoon? If yes, that is a €300 job done in minutes. If not, it stays a nice video.
Source: Google DeepMind
If it doesn’t save you time or money, you won’t see it here
- Facts, not hypeEvery price and number is checked against the official source
- Always the real priceYou see what it costs before you click
- I say when it’s weakEven when it’s OpenAI or Google
