🔍 Read the full analysis: What It’s Like To Walk Into A World Built By Google’s Genie 4 on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Google DeepMind demonstrated Genie 4 on Oct. 27, describing it as a real-time world model that generates interactive scenes with sound and responds to text and user controls. The demonstrations are company-selected, public access has not opened, and DeepMind says scenes can drift after a few minutes.
Google DeepMind demonstrated Genie 4 on October 27, showing an AI model that generates a scene users can move through and alter with text while it runs. The company says it can create matching audio and more realistic people and physics, but the system is not yet broadly available, its demonstrations were selected by DeepMind, and the company says scenes may drift during longer sessions.
In DeepMind’s demonstration, a user enters a scene description and then moves through the generated environment with keyboard and mouse controls. Text instructions can change conditions or add elements while the scene continues. DeepMind also says the model generates ambient sound, music and dialogue alongside the visuals. These capabilities are the company’s descriptions; the announcement did not provide benchmark figures for image quality, interaction speed or consistency.
DeepMind says Gemini 3.8 Flash works alongside Genie 4 to interpret goals, break them into steps and monitor generated frames. The company describes this as a way to help keep scenes coherent. Its robotics example showed a simulated Apptronik Apollo humanoid playing checkers; according to the source material, no physical robot took part. The example is a simulation, not evidence of a robot completing the task in the real world.
Access is currently limited to trusted testers. DeepMind says it plans to offer Genie 4 through Project Genie to Google AI Ultra subscribers and to early partners using the Gemini API, but it has not announced a launch date or API price. Reports cited in the source describe some scenes lasting about two minutes; that duration is not stated in DeepMind’s own announcement.
Infographic 01 · How Genie grew up
From flat sketches to worlds with sound, in under three years.
Each step as Google DeepMind described it. Session limits and consistency are different measures, so read the labels.
- Feb 2024
Genie 1Simple 2D worlds11 billion parameters, learned from unlabelled internet videos - Dec 2024
Genie 23D worlds from a single imageConsistent “for up to a minute”; real time only as a slimmed-down, lower-quality version - Aug 2025
Genie 3First real-time Genie720p at 24 frames per second, consistent “for a few minutes”, memory about one minute back - Jan 2026
Project GenieUS AI Ultra subscribers can play60 seconds per world, partly because of compute costs - May 2026Project Genie goes globalIn the $200 AI Ultra tier, for users 18 and older
- 27 Oct 2026
Genie 4Sound, live steering, more realismResearch preview for trusted testers; Ultra subscribers and API partners “coming soon”
Genie 3 vs. Genie 4
As described by Google DeepMind. Genie 4 column = vendor claims unless marked.
| Feature | Genie 3 (Aug 2025) | Genie 4 (Oct 2026) |
|---|---|---|
| Sound | Genie 3: None | Genie 4: Generated with the picture: ambient noise, music, dialogue |
| Steering | Genie 3: Keys, plus “promptable world events” | Genie 4: Free-form text changes while it runs; Gemini adapts and checks them |
| Look | Genie 3: 720p, 24 fps | Genie 4: “Unprecedented realism”; no resolution or frame rate published |
| How long | Genie 3: “A few minutes”; Project Genie capped at 60 s | Genie 4: “A few minutes” (DeepMind); up to ~2 min consistent scene reported |
| Access | Genie 3: Small preview, then Project Genie | Genie 4: Trusted testers; Ultra and API partners “coming soon” |
Sources: Genie paper, arXiv, 23 Feb 2024; Genie 2 blog, 4 Dec 2024; Genie 3 blog, 5 Aug 2025; TechCrunch, 29 Jan 2026; Google, 19 May 2026; Google DeepMind, Genie 4 blog, 27 Oct 2026; Genie 4 model page (limitations, disclaimer); TechCrunch, 27 Oct 2026. Table wording: Thorsten Meyer.
Powered by Thorsten Meyer AI
Interactive Worlds Beyond Prebuilt Games
Genie 4 points to a different approach to generating digital environments: rather than render a world from a stored 3D map, the model predicts how the scene should look as a person moves and gives instructions. If that approach becomes reliable, it could let developers sketch or test environments without first building every object and rule in a conventional game engine. DeepMind says testers have explored uses in game prototyping and film previsualization, but those are reported feedback, not established commercial results.
The same technology could matter for research that needs simulated environments. DeepMind’s robotics example shows one possible direction, while remaining a computer simulation. Separately, the source reports that Waymo uses Genie 3, a predecessor, in its driving simulator. That does not establish that Genie 4 is being used for autonomous-vehicle training. For ordinary viewers, the combination of generated video and sound also increases the importance of clear provenance: DeepMind says Genie 4 outputs carry an invisible SynthID watermark.
Popular right now
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Black
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Black
As an affiliate, we earn on qualifying purchases.
From Genie 3 to Real-Time Scenes
Genie is DeepMind’s series of models for generating environments that respond to user input. The initial Genie, announced in February 2024, produced simple 2D worlds. Genie 2 followed in December 2024 with 3D scenes generated from a single image, but its higher-quality operation was not real time. DeepMind presented Genie 3 in August 2025 as a real-time model capable of 720p video at 24 frames per second and consistency for a few minutes, according to the source material.
Google later made Genie 3 available to paying US subscribers through Project Genie, with a reported 60-second limit per world. Genie 4’s announced additions include audio generation, text-led changes during interaction and scenes DeepMind says have more natural physics and people. The evidence available so far is the company’s demonstration material and description; independent performance comparisons are not provided in the supplied source.
“Details and objects “can shift or drift over time.””
— Google DeepMind
Access, Duration and Reliability
Several practical details remain open. DeepMind has not announced when Project Genie access will begin, which countries will be included, or how much Gemini API use will cost. The company has not published benchmark results that would allow readers to compare Genie 4’s quality or reliability with other systems.
There is also a gap between curated demonstrations and sustained use. DeepMind warns that objects and details may drift, and says interaction lasts a few minutes. The source material cites reports of scenes lasting about two minutes, but the exact duration may depend on the setup and is not specified in the company’s announcement. It is not yet clear how well the model handles longer sessions, repeated changes, or complex environments outside selected demos.
Potential applications also need to be treated cautiously. Film and game uses are attributed to tester feedback, while the humanoid example was simulated. The source does not establish that Genie 4 is being deployed commercially for these purposes, or that its generated worlds are suitable for training real-world robots.
Google’s Planned Tester Access
The next stated step is access for Google AI Ultra subscribers through Project Genie and early partners through the Gemini API. DeepMind has not provided a date, so broader availability cannot be timed from the announcement. Its rollout and any further technical documentation may clarify how long sessions can run, what controls users have and how developers can integrate the model.
Until then, Genie 4’s demonstrated capabilities should be read as a company presentation rather than a public product test. Independent evaluations, access terms and performance details will be needed to judge whether the model can support the uses DeepMind and its testers describe.
Key Questions
What is Google Genie 4?
Genie 4 is a world model from Google DeepMind. The company says it generates scenes frame by frame as a user moves or gives instructions, rather than drawing from a stored 3D map like a conventional game engine.
Can the public use Genie 4 now?
No. DeepMind says access is currently for trusted testers. It plans to offer the model to Google AI Ultra subscribers through Project Genie and to early Gemini API partners, but has not announced a date or API price.
Does Genie 4 generate sound?
DeepMind says it generates audio alongside visuals, including ambient sound, music and dialogue. This is the company’s description of the model; the supplied material includes no independent benchmark of audio quality.
How long can a Genie 4 scene last?
DeepMind says the model supports a few minutes of continuous interaction and warns that details can drift. Reports cited in the source put some scenes at about two minutes, but that figure is not in DeepMind’s own announcement.
Does the robotics demo show a real robot using Genie 4?
No. The cited demonstration showed a simulated Apptronik Apollo humanoid playing checkers. The source material says no physical robot was involved, so the clip does not establish real-world robot performance.
Source: ThorstenMeyerAI.com
