🔍 Read the full analysis: Can You Explore Google’s Genie 4 Worlds? What To Know on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Google DeepMind introduced Genie 4 on October 27, 2026, presenting it as a real-time AI model that generates interactive worlds from images, movement and text instructions. Public access is not available: the company says trusted testers can use it, with access for Google AI Ultra subscribers and Gemini API partners planned for later. DeepMind says sessions last a few minutes and that scenes can drift; it has not published benchmark results or a release date.
Google DeepMind demonstrated Genie 4 on October 27, 2026, describing it as a model that generates an interactive scene as a user moves through it and gives instructions. The company says it can produce real-time visuals and audio, but the system is not open for public use: access is currently limited to trusted testers, and DeepMind has not announced a wider-release date.
In DeepMind’s demonstrations, users steer scenes with keyboard and mouse controls and can type instructions to change conditions, such as making it rain or adding an object. The model generates the next frames in response, rather than displaying a prerecorded clip. DeepMind also says Genie 4 generates sound alongside the imagery, including environmental audio, music and dialogue. These capabilities are based on company-selected demonstrations; DeepMind has not released independent test results or benchmark figures.
DeepMind describes Gemini 3.8 Flash as an agent that can interpret a user’s goal, break it into steps and monitor the generated scene. Its announcement also includes a robotics simulation involving an Apptronik Apollo humanoid playing checkers. That was a simulated scenario, not a demonstration of a physical robot carrying out the task. DeepMind separately says film previsualization and game prototyping are uses testers have explored; these are early reports, not established commercial deployments.
The company says the generated worlds support a few minutes of continuous interaction, while warning that objects and details may shift or drift over time. TechCrunch and Bloomberg, as cited by German outlet heise, have reported scenes lasting about two minutes, but that figure does not appear in DeepMind’s own description. DeepMind says future access is planned for Google AI Ultra subscribers through Project Genie and for early Gemini API partners. It has not provided a launch date or API pricing.
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
Why Interactive Worlds Matter
Genie 4 points toward a different way of creating and exploring digital scenes: instead of building a complete game environment in advance or rendering a fixed video, the model generates what appears next in response to a user. If the approach becomes reliable and affordable, it could help creators test visual ideas, sketch game concepts or simulate situations without first constructing a full 3D environment. Those are potential uses, not proven outcomes at this stage.
The gap between the demonstrations and public availability matters to anyone hoping to try the system. There is no open consumer release, and no published price or performance testing to show how it would work at scale. For companies, the announcement is a sign to watch rather than evidence that a new production tool is ready to replace established workflows. For viewers, generated scenes with synchronized sound may also make synthetic footage more convincing. DeepMind says its Genie 4 video and audio outputs carry an invisible SynthID watermark, though the announcement does not establish how reliably that watermark can be detected after content is edited or shared.
Products Worth Considering
Samsung 27-inch Odyssey OLED G6 QD-OLED QHD Gaming Monitor
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
As an affiliate, we earn on qualifying purchases.
From Short Clips to Live Worlds
Genie is part of a series of DeepMind research systems. The first Genie, introduced in February 2024, generated simple two-dimensional environments from video training data. Genie 2, announced in December 2024, created three-dimensional worlds from a single image and maintained consistency for up to a minute, according to DeepMind. Its real-time demonstrations used a reduced-quality version.
In August 2025, DeepMind introduced Genie 3 as its first real-time world model, describing output at 720p and 24 frames per second with consistency lasting a few minutes. The company later made Genie 3 available to paying U.S. subscribers through Project Genie in January 2026. TechCrunch reported a 60-second limit per world for that service and linked the limit partly to computing costs. DeepMind has not said whether Genie 4 will have the same access limits or costs.
Genie 4’s central difference in the announcement is its combination of interactive scene generation, audio and live text instructions. DeepMind also says it uses Gemini Omni to anchor simulations in real-world appearance and behavior. The company’s claims about realism and physical behavior remain vendor claims until they can be tested independently.
“The model supports “a few minutes of continuous interaction.””
— Google DeepMind
Access, Limits and Testing
Several practical details remain unknown. DeepMind has not given a public launch date, confirmed the final availability terms for Project Genie or the Gemini API, or announced API pricing. It has also not published independent benchmark results that would let users compare Genie 4’s speed, realism or consistency with other systems. The demonstrations show selected examples, not a complete account of how the model performs across different prompts and settings.
The model’s stated ability to maintain a scene for a few minutes does not mean that every object or location stays consistent throughout that period. DeepMind explicitly warns about drift, and it has not described how the system’s memory behaves in every use case. The company has also not provided detailed validation of the SynthID watermark’s detection under common editing and redistribution conditions. Media reports of roughly two-minute scenes offer additional context, but are not a substitute for a full public specification.
Awaiting Public Access Details
For now, Genie 4 remains in a trusted-testing phase. DeepMind says access is expected to come to Google AI Ultra subscribers through Project Genie and to early partners using the Gemini API, but the company has not said when either group will receive it. Interested users should treat “coming soon” as an unspecified timeline, not a confirmed release date.
The next useful developments will be a public access announcement, clearer pricing and usage limits, and independent evaluations of how well scenes remain coherent during extended interaction. Details about the model’s computing requirements and the handling of generated audio and video would also help creators and businesses judge whether it fits their work. Until those details are available, Genie 4 is best understood as a research and product demonstration with promising capabilities and acknowledged limits.
Key Questions
Can the public try Google Genie 4 now?
No. DeepMind says Genie 4 is currently available only to trusted testers. The company says access is planned for Google AI Ultra subscribers through Project Genie and early Gemini API partners, but it has not set a public release date.
What can Genie 4 do in the demonstrations?
Users can move through generated scenes with keyboard and mouse controls and type instructions to change what happens. DeepMind says the system generates visuals and audio in real time, including environmental sounds, music and dialogue.
How long can users explore a Genie 4 world?
DeepMind says the model supports a few minutes of continuous interaction, while warning that objects and details can drift. Some media reports cite scenes lasting about two minutes, but DeepMind has not stated that figure in its own description.
Is the robot demonstration a real-world test?
No. The Apollo humanoid playing checkers was shown in a robotics simulation. The demonstration did not show a physical Apptronik robot performing the task.
Has Google published independent performance results?
DeepMind has not published benchmark figures or independent evaluations in the supplied announcement material. Its claims about realism, physics and other capabilities are company descriptions of the system, illustrated with selected demonstrations.
Source: ThorstenMeyerAI.com
