🔍 Read the full analysis: What You Should Know Before Exploring Genie 4’S AI Worlds 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 presented Genie 4 on Oct. 27, describing an AI model that generates interactive visual worlds with sound and accepts live text instructions. The demonstrations are company-selected, public access has not opened, and DeepMind says scenes may drift and support only a few minutes of interaction.
Google DeepMind demonstrated Genie 4 on October 27, presenting a model that generates interactive visual worlds, sound and responses to typed instructions as users move through a scene. The announcement matters because it points toward AI-generated environments that can be explored rather than simply watched, but public access is not yet available and the demonstrations do not establish how well the system performs beyond selected clips.
DeepMind describes Genie 4 as a world model: it predicts and generates what happens next from prior frames and user inputs, rather than drawing a view from a stored 3D map like a conventional game engine. In the company’s demonstrations, users move with keyboard or mouse controls and type changes such as making it rain or adding an object. DeepMind says the system can generate ambient sound, music and dialogue alongside the visuals.
The company also says Gemini 3.8 Flash can interpret a user’s goal, break it into steps and monitor the generated scene between inputs. DeepMind describes improvements in visual realism, people and physics, including simulations of collisions and aerodynamics. Those descriptions are vendor claims: the announcement material cited here provides no benchmark figures, and the clips were selected by DeepMind.
DeepMind says Genie 4 is currently available only to trusted testers. It says access is coming soon for Google AI Ultra subscribers through Project Genie and for early Gemini API partners, but has not announced a date or API price. The source reports describe the proposed subscriber tier as costing $200 a month. DeepMind says interaction can continue for a few minutes before limits become apparent; TechCrunch and Bloomberg, as cited by heise, reported clips of up to about two minutes, a figure 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
From Generated Clips to Interactive Worlds
Genie 4’s significance lies in the shift from generating a finished video to creating a scene that responds as a person explores it. If the system can keep environments coherent while reacting to movement and text, it could help teams make game prototypes, plan film scenes or test ideas without first building a full digital environment. DeepMind says testers have reported film previsualization and game prototyping as uses; those are early feedback, not measured outcomes.
The announcement also has implications for robotics research, but the evidence needs careful framing. DeepMind showed a simulation involving Apptronik’s Apollo humanoid playing checkers, with a separate Gemini model planning task steps. According to the source’s account, no physical robot took part, so the clip demonstrates a simulated task, not a robot’s real-world performance. Alphabet’s Waymo already uses the predecessor Genie 3 in its driving simulator, according to the source, showing a research and simulation path without proving Genie 4 is ready for commercial use.
More convincing generated audio and images could also make synthetic media harder for viewers to identify. DeepMind says Genie 4’s video and audio outputs carry an invisible SynthID watermark. The announcement does not establish how reliably that watermark can be detected or preserved after editing, so it should not be treated as a complete answer to authenticity concerns.
Products Worth Considering
Samsung 32-Inch Odyssey OLED G8 (G80SH) 4K Gaming Monitor
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
How Genie’s Capabilities Have Evolved
The Genie project has advanced through several research versions. The first model, introduced in February 2024, generated simple 2D worlds. Genie 2, announced in December 2024, created 3D environments from a single image and maintained consistency for up to a minute, though real-time use required a reduced-quality version. In August 2025, DeepMind described Genie 3 as its first real-time world model, running at 720p and 24 frames per second and sustaining a scene for a few minutes.
Google later made Genie 3 available to paying US subscribers through Project Genie, according to the source. TechCrunch reported a 60-second limit per world at that time, partly linked to computing costs. Genie 4’s stated additions are native audio generation, live text-driven changes and the use of a Gemini model to guide interactions. The predecessor’s rollout offers relevant context, but it does not confirm Genie 4’s eventual price, access rules or runtime limits.
“Details and objects can shift or drift over time.”
— Google DeepMind, describing Genie 4’s runtime limits
Access, Runtime and Reliability
Several practical details remain unsettled. DeepMind has not published a public release date, API pricing or benchmark results in the material described here. It is also unclear what access restrictions, geographic limits or usage caps will apply when Project Genie access expands. The reported $200 monthly price is associated with the Google AI Ultra tier, not a confirmed Genie 4-specific charge.
How well the generated worlds hold together beyond short demonstrations is also unknown. DeepMind acknowledges that objects and details may drift, and its stated interaction period is only a few minutes. Because the announcement relies on company-selected demonstrations, readers cannot yet assess typical output quality, failure rates or performance across a wide range of prompts. The claims about film, game and robotics use have not been accompanied here by independent tests or quantified results.
What to Watch in the Rollout
The next milestone is DeepMind’s planned expansion to Google AI Ultra subscribers through Project Genie and to early Gemini API partners. The company has not given a date, so availability remains a stated plan rather than a scheduled launch. Further details would clarify who can use the model, how much access costs and whether users can save or share generated worlds.
For a clearer assessment of the technology, readers will need to see independent testing and fuller technical information on runtime, consistency, audio generation and watermark detection. Until then, Genie 4 is best understood as a research-stage demonstration with limited tester access, rather than a broadly available tool with established performance.
Key Questions
Can the public use Genie 4 now?
No. DeepMind says access is limited to trusted testers for now. It has described future access for Google AI Ultra subscribers through Project Genie and early Gemini API partners, but has not announced a release date.
What does Genie 4 do?
DeepMind describes it as a world model that generates a scene as a user moves through it and gives instructions. The company says it can produce visuals and sound together, with a Gemini model helping interpret goals and guide the scene.
How long can a Genie 4 world run?
DeepMind says it supports a few minutes of continuous interaction and warns that details may drift. TechCrunch and Bloomberg, as cited by heise, reported scenes of about two minutes; that specific figure was not in DeepMind’s own text.
Has Genie 4 been shown controlling a real robot?
The robotics example described in the source was a simulation involving Apptronik’s Apollo, not a physical robot performing the task. The demonstration should not be read as evidence of real-world robot capability.
Does Genie 4 mark the first AI system to generate audio?
No such industry-wide claim is established. DeepMind calls it its first real-time world model from the company to generate audio natively; the source notes that another system, Runway’s GWM Worlds 2, also generates audio.
Source: ThorstenMeyerAI.com
