Newshot

Epic Evolution! Google's Genie 4 World Model Opens for Hands-On Experience: Change the Weather and Summon Characters While Playing, Everything You See Is Interactive

AI summary

Google has launched a new Project Genie experience for eligible Google AI Ultra subscribers, offering faster, more creative, and more social virtual worlds. The update opens access to Genie 4 for the first time, bringing upgrades to Genie 3, a Project Genie agent that helps users realize their ideas, and a community gallery for discovering, exploring, and remixing worlds. Genie 4 lets users modify already-created worlds by entering prompts mid-session, such as switching the weather, adding characters, or changing the direction of the action, and it adds real-time native audio for the first time.

Translation(model translation, not the source's own text)

Epic Evolution! Google's Genie 4 World Model Opens for Experience: Edit Weather Mid-Play, Summon Characters, Everything You See Is Interactive

IT之家 November 19 news: Earlier this year, Google launched an experimental research prototype, Project Genie, which lets users generate and explore interactive virtual worlds. Early testers fed back that they wanted stronger control over what happens within the worlds, along with more convenient ways to build and share.

Now, Google is rolling out a new Project Genie experience to eligible Google AI Ultra subscribers, making these virtual worlds significantly faster to respond, more creative, and more social.

This update opens access to Genie 4 for the first time, bringing a new upgrade over Genie 3, the Project Genie agent that helps users bring their ideas to life, and a community gallery for discovering, exploring, and remixing worlds.

Project Genie combines Google's most advanced world models with the Gemini family of models, letting users generate, enter, and shape all kinds of virtual worlds directly in the browser.

Genie 4 brings a whole new world simulation experience: as Google's most realistic and most interactive world model to date, Genie 4 lets users enter prompts mid-session to modify worlds they have already created.

By typing a new prompt directly while navigating, users can edit the environment in real time — switching the weather, adding characters, and changing the direction of the action are all possible. At the same time, Genie 4 for the first time pairs visual content with real-time native audio that adapts dynamically as the scene progresses, covering everything from ambient sound effects to character dialogue.

Genie 3 gets a number of capability upgrades: Genie 4 offers the highest-fidelity visuals, physics, and native audio, while Genie 3 remains in the model selector, suited to scenarios that need longer generation times and spaces generated from panoramas.

This Genie 3 update includes: extended session length and world chaining, raising the explorable duration of a single generation from 60 seconds to up to 3 minutes, and allowing multiple worlds to be chained into a longer journey; support for 360-degree panoramas as a generation basis, so new worlds can be generated from a panorama seed; and support for real-time in-world events, allowing the environment to be changed dynamically through natural language prompts during exploration.

Project Genie 3 can use real-time triggerable events to chain different worlds — forests, underwater ruins, frozen seas, and more — into one continuous flight journey; it can also use 360-degree panorama grounding to keep the Petra canyon and ancient building facades consistent as the view rotates; and it can combine long-duration generation, panorama bases, and real-time events to change the world at any point during a flight, even summoning a dragon as a mount.

The new intelligent agent lowers the barrier to creation: the new Project Genie agent expands creation beyond single text or image prompts, making it as simple as a conversation to build an interactive universe. Users only need to describe their idea in natural language, and the agent will orchestrate Google's suite of multimodal models to generate the initial visuals that set the world's style, along with custom background music that matches the mood.

Users can also add Gemini models to the session during exploration: it observes what is happening on screen, pops up contextually relevant suggested options, or automatically triggers in-world events as the journey progresses. For example, turning a natural language prompt into an interactive spacewalk, summoning an alien spacecraft on command, or transforming a hillside café into a desert glass pavilion. Once creation is complete, users can publish their custom worlds to the gallery for others to explore and remix.

The new gallery brings a wealth of content to explore: when users just want to jump straight into exploring, the new Project Genie gallery brings together a large collection of curated worlds and community creations. Users can turn personal photos into explorable spaces in "Living Memory," roam 360-degree panorama-based prehistoric habitats in "Dino Field Guide," or have Gemini serve as a real-time referee in "Rock Paper Scissors," pitting themed objects on a table against each other.

The gallery also showcases works created with pioneering creators who gained early access to the tools: "Villa Giallo," created with creative technologist Will (aka BLVCKL!GHT), uses Genie 4 to transform a Gothic manor into a room-by-room interactive mystery space; and "The Life of Ants," created with filmmaker Anand Gandhi, leverages Genie 4's photorealistic quality to let users explore a complex ant colony from an ant's perspective.

Beyond exploring curated worlds, users can also kick off their own creations with a simple prompt, or remix existing gallery works together with the Project Genie agent — adjusting the underlying world structure, soundtrack, or rules to create their own content. Once creation is complete, it can be published to the community gallery for others to explore and remix.

Google says it always develops and promotes interactive world models responsibly. All video and audio generated in Project Genie carries SynthID watermarks, and all model inputs and outputs go through safety screening.

Genie 3 and Genie 4 remain world models at the frontier research stage. As research continues, Google is actively working to address core challenges such as camera latency, complex physical interactions, real-time audio accuracy, and long-term scene persistence.

Google Genie 4 official promotional image
Google Genie 4 official promotional image

Starting today, eligible Google AI Ultra subscribers can experience the new Project Genie, and Google AI Pro members can join the waitlist on the same page.

Related Articles

  • Google signs over $500 million clean energy power purchase agreement: to purchase 50MW of deep geothermal energy from Quaise starting in 2030
  • Google announces its 2026 US Best of the Year list: "Arknights: Endfield" takes Best Game of the Year
  • Google Wallet launches on Apple's iOS system, contactless payments not yet supported
  • Microsoft and Apple suffer successive outages, Google responds to Gmail outage rumors: "service operating normally"
  • Google updates Gemini privacy controls, allowing chat history retention without AI training participation
  • Google Gemini 4 Argon safety test goes awry: 40 AI agents attack real websites without authorization and share vulnerabilities with each other

The translation is machine-generated and individual terms may be off. Nothing is rewritten or cut; the original is authoritative.

Published by IT之家
This site only summarises and aggregates; copyright stays with the original author. Headlines and summaries follow the source — where they differ, the original wins.
Read the original ↗