Google DeepMind has launched Genie 4, a significant upgrade to its artificial intelligence platform that constructs virtual environments. This model was released as a research preview on October 2026, marking a departure from traditional AI video generation. Instead of merely producing standard videos, Genie 4 creates photorealistic, three-dimensional spaces that users can actively explore and modify during the simulation.

Real-Time Modifications and Audio Integration

One of the standout features of Genie 4 is its ability to allow real-time edits. Users can change elements such as camera angles, weather conditions, or even introduce new characters without pausing the simulation. This functionality enhances user engagement by providing a seamless experience. Furthermore, Genie 4 incorporates live audio capabilities, generating sound that corresponds to the visual elements on screen. This includes ambient noise, music, and even dialogue from digital characters that respond to user inputs.

The duration of simulations has also been extended, now running for up to two minutes, which is double the time offered by the previous version, Genie 3. Google has indicated improvements in real-world physics and the realism of character animations, making the virtual environments more immersive.

To facilitate these advancements, Genie 4 collaborates with Google’s Gemini AI. Gemini functions as a director, orchestrating the logic and sequence of events while Genie focuses on rendering the visuals. Tejas Kulkarni, a research director at Google, highlighted that the future of AI will likely involve such hybrid systems, where Gemini generates the text or code and Genie executes the simulation in real time.

Applications and Market Competition

The market for AI simulations is becoming increasingly competitive, with companies like Runway launching their own products, such as GWM Worlds 2. Major corporations, including AMD, Amazon, and Rivian, are acquiring smaller AI startups to enhance their technological capabilities. Google envisions Genie 4 being utilised across various sectors, including gaming, interactive education, and robotics. Diego Rivas, an engineer at Google DeepMind, also noted its potential as a valuable resource for filmmakers, enabling rapid storyboarding and allowing creators to maintain their creative momentum without lengthy video rendering times.

Currently, Google offers Veo for high-definition video clips, and there are plans for developers to eventually leverage Gemini Omni, Google’s multimodal model, to convert Genie 4 prototypes into final creative products. This integration could streamline the creative process across multiple platforms.

Current Limitations and Accessibility

Despite its innovative features, Genie 4 is still in the early stages of development. Rivas acknowledged that prolonged simulations may lead to issues such as “drifting,” where objects lose detail or vanish altogether. The team is also focused on improving audio synchronisation and ensuring that text remains legible within the generated environments.

At present, Genie 4 is accessible only to a select group of trusted testers. In the upcoming weeks, it will be made available to Google AI Ultra subscribers through Project Genie, which is priced at Rs 8,300 per month. Additionally, early access will be granted to partners via the Gemini API and Google AI Studio, expanding the reach of this cutting-edge technology.

As Google continues to refine Genie 4, its implications for various industries could be substantial, paving the way for new forms of interactive experiences and creative opportunities.

Long or Short, get news the way you like. No ads. No redirections. Download Newspin and Stay Alert, The CSR Journal Mobile app, for fast, crisp, clean updates!

App Store – https://apps.apple.com/in/app/newspin/id6746449540

Google Play Store – https://play.google.com/store/apps/details?id=com.inventifweb.newspin&pcampaignid=web_share