🔍 Read the full analysis: Google’s Genie 4 Explained: Its World-Building Features And Limits on ThorstenMeyerAI.com
Get privacy and security gear 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 presented Genie 4 on October 27, 2026, describing a real-time model that generates interactive scenes with audio and responds to text instructions. The company has not published benchmark results, says access will initially be limited, and acknowledges that scenes can drift during sessions lasting a few minutes.
Google DeepMind demonstrated Genie 4 on October 27, presenting it as a real-time world model that generates interactive scenes, sound and changes prompted by text. The system is not publicly available: DeepMind says it is being tested by trusted users and will later reach some Google AI Ultra subscribers and Gemini API partners, while acknowledging that scenes can shift or drift over time.
A world model generates what happens next from the scene so far and from a user’s input, such as a keyboard command or typed instruction. Unlike a conventional game engine, which renders views from a stored map of a world, Genie 4 predicts successive frames. DeepMind says users can move through a scene and ask for changes, including different weather, camera movement or new objects, while it is running.
The company describes four main capabilities: interactive controls, more realistic physics and people, audio generated alongside video, and Gemini working as a guiding agent. DeepMind says the audio can include environmental sounds, music and dialogue. It also says Gemini 3.8 Flash can interpret goals, break them into steps and monitor generated frames. These are vendor descriptions illustrated with selected company demos; DeepMind has not released benchmark figures in the material provided.
Access remains restricted to trusted testers. DeepMind says Genie 4 is coming to Google AI Ultra subscribers through Project Genie and to early Gemini API partners, but has announced no launch date or API price. The source reports Google’s current Ultra tier at $200 a month. DeepMind also says Genie 4 video and audio outputs carry an invisible SynthID watermark.
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
Uses and Limits of Generated Worlds
Interactive generated environments could let developers test ideas before building full game levels or filming sets. DeepMind says testers have given feedback on film previsualization and game prototypes. That is early user feedback, not evidence that Genie 4 is ready to replace established production tools or game engines.
The robotics angle is also relevant but needs careful qualification. DeepMind showed a research simulation of Apptronik’s Apollo humanoid playing checkers, with a model breaking the task into steps. The cited account says no physical robot took part. Such demonstrations may help researchers explore simulated tasks, but they do not establish that a robot can perform them in the real world.
For audiences and publishers, systems that create video and matching sound raise questions about identifying synthetic media. Google’s stated use of SynthID watermarking is one response, but the material does not establish how reliably the watermark can be detected or preserved when content is edited or shared. The system’s practical value will also depend on how long it can maintain a stable world and how much access costs.
Products Worth Considering
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From Genie 3 to Genie 4
Google’s Genie line has developed from simple generated environments toward real-time interaction. The source account says the original Genie, introduced in February 2024, generated simple 2D worlds. Genie 2 followed in December 2024, creating 3D worlds from a single image and maintaining them for up to a minute, though real-time operation required a lower-quality version.
Genie 3, presented in August 2025, was described by DeepMind as its first real-time world model, running at 720p and 24 frames per second and remaining consistent for a few minutes. Google later made it available to paying US subscribers through Project Genie in January 2026. TechCrunch reported a 60-second limit per world, citing compute costs. Those figures concern Genie 3 and should not be treated as independent performance measurements of Genie 4.
Genie 4’s headline changes, as described by DeepMind, are native audio generation, live text-based scene changes and more realistic movement and people. Audio generation is not unique across the industry: TechCrunch noted that Runway’s GWM Worlds 2 also generates audio. The demos establish what Google chose to show, not a comparative assessment against other systems.
“Details and objects can shift or drift over time.”
— Google DeepMind
Performance and Access Still Unclear
DeepMind has not provided benchmark numbers in the supplied material, and the public demonstrations were selected by the company. It is not yet clear how Genie 4 performs across a broad range of prompts, how consistently it follows instructions, or how its results compare with rival systems under equivalent tests.
The reported duration also needs qualification. DeepMind says interaction can continue for a few minutes; TechCrunch and Bloomberg, as cited by heise, report scenes lasting about two minutes. That more specific figure is not in DeepMind’s own stated description. The exact session limit, resolution and performance available to future subscribers have not been confirmed here.
Google has not given a release date, API pricing or full availability details. It is also unclear how SynthID performs after content is edited or redistributed. The robotics demonstration remains a simulation, and no evidence in the source material shows Genie 4 controlling a real robot.
What Google Has Yet to Announce
The next stated step is access for Google AI Ultra subscribers through Project Genie and early Gemini API partners. DeepMind has not set dates for either group or provided API prices, so availability should be treated as planned rather than scheduled.
Further information will be needed to judge the model beyond its presentation: independent tests, clear session and quality specifications, pricing, and details on how the system handles longer interactions. Until those are available, Genie 4 is best understood as a demonstrated research and product direction, not a broadly accessible tool with independently established performance.
Key Questions
What is Google Genie 4?
Genie 4 is a world model that generates the next parts of an interactive scene in response to movement and other user input, rather than rendering a world from a conventional stored 3D map.
Can the public use Genie 4 now?
No public release is described. Google DeepMind says it is available to trusted testers and is planned for Google AI Ultra subscribers through Project Genie and early Gemini API partners. No date has been announced.
What features did DeepMind show?
The company demonstrated keyboard and mouse movement, text-directed scene changes, generated audio and a Gemini model helping guide tasks. Claims about realism and physics come from DeepMind’s descriptions and selected demos; benchmark figures were not provided.
How long can a Genie 4 scene run?
DeepMind describes a few minutes of continuous interaction and warns that details can drift. Reports of scenes lasting about two minutes come from TechCrunch and Bloomberg as cited by heise, not from the company’s stated wording.
Does the robot demonstration show a real Apollo robot?
No. The described Apollo checkers demonstration was a simulation; the source account says no physical robot was involved.
Source: ThorstenMeyerAI.com
