Google DeepMind has a new world model, and the gaming industry is probably checking its portfolio again.
Genie 4 improves on Genie 3 in almost every way, according to Ars Technica’s reporting. The upgrades in focus are audio and more realistic simulations.
What Genie 4 changes
A world model generates a virtual environment on the fly instead of loading pre-built scenery. Nothing is rendered ahead of time. The system produces what you see in response to what you do.
Genie 3 already worked this way, with clear limits. It launched in August 2025 with real-time interactive output at 720p and 24 frames per second. It also had a one-minute memory limit, so the world it built could start losing track of itself fairly quickly.
Genie 4 is aimed at those rough edges. DeepMind researchers had previously discussed better realism, longer memory and audio as features for future versions of the model. Sound and realism are the improvements now attached to the new release.
Realism has also been climbing on the Genie 3 side. As of late October 2026, Project Genie integrated Google Street View data to produce photorealistic street-level environments, with a broader rollout planned.
How we got here
The Genie line has moved fast. Genie 1 arrived in February 2024. Genie 2 followed in December 2024, adding 3D support and higher resolution.
Genie 3 was the version that escaped the lab. On January 29, 2026, Google launched Project Genie, an experimental platform giving limited Genie 3 access to Google AI Ultra subscribers in the US. Users could sketch and explore virtual worlds in short sessions, capped at a single minute due to compute constraints.
Wall Street did not need more than a minute. Following the January 2026 rollout, Take-Two shares fell about 8-10%, Roblox dropped about 12-13%, and Unity slid about 21-24%.
DeepMind says this isn’t about games
DeepMind frames it differently. Google positions world models as a key ingredient in the path toward artificial general intelligence, or AGI. The idea is that an AI able to simulate a consistent world, with working physics, has a better shot at understanding the real one.
Genie 3 was already pitched for uses well beyond entertainment. Those include robotics, simulation, animation and historical exploration.
What this means
For gaming investors, the obvious question is whether Genie 4 triggers a repeat of January. Unity’s roughly 21-24% decline stands out because engine tools sit closest to what a world model replaces.
The counterargument is that limits still apply. Genie 3 needed a one-minute cap because generating worlds in real time eats huge amounts of compute. Unless Genie 4 dramatically changes that math, it remains a premium, rationed experience rather than a replacement for a finished commercial game.
For Google, Genie 4 strengthens the case for its most expensive AI subscription tier. Paid early access turns research into a product that also generates user feedback. That loop of shipping, learning and iterating is how the series went from Genie 1 to Genie 4 in a little over two years.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.