October 27, 2026 · Models
Raia Hadsell, Diego Rivas, and Hang Qi

Today, we’re officially introducing Genie 4, our new world model that delivers unprecedented realism and significantly enhances real-time interactive simulations. Genie 4 is natively multimodal, generating synchronized audio alongside the worlds it builds—including music, ambient soundscapes, and interactive dialogue.
The Genie 4 model integrates and builds upon the capabilities of the Gemini model family: Gemini Omni gives simulated worlds their realistic look and behavior, while other Gemini models provide the complex reasoning required to maintain simulation coherence and continuity.
With this foundation, all you need to do is trigger an event, and Genie 4 simulates what unfolds next: how the world responds, what sounds it makes, and how the people within it act, react, and converse.
Understanding and simulating the physical world is a crucial milestone on the path to artificial general intelligence (AGI). Genie 4 substantially elevates our ability to generate safe, infinitely variable simulated environments—powering training environments for robotics and autonomous agents while unlocking entirely new forms of interactive experiences.
We have begun rolling out a research preview of Genie 4 to trusted testers to gather feedback ahead of a broader release to consumers and developers. Initial access is available to Google AI Ultra subscribers within Project Genie, as well as early access partners across the Gemini API and Google AI Studio. Join the waitlist for Genie 4 in Project Genie here.


Genie 4 delivers superior photorealism, fluid motion, and native audio generation. All accompanying audio heard in the video above was generated directly by Genie 4.
Four core breakthrough capabilities of Genie 4
Below are real-time interactive screen recordings from Genie 4, where users navigate perspective and camera movements using standard keyboard controls while steering scene dynamics with text prompts.
Reshaping world simulations in real time
With Genie 4, you can modify simulated scenes in real time using natural language—whether shifting the weather, reframing the camera, or introducing new characters and objects. Gemini dynamically adapts your prompts in context to blend seamlessly with the surrounding environment. Between each of your inputs, Gemini continuously monitors and directs the scene, preserving the continuity of the unfolding world.
Turn on audio to hear the sounds of the world generated by Genie 4.
Slide 1 of 3


Prompt: Use the reference image as the exact first frame and visual baseline for the scene. Retain the identical floating rectangular platform, weathered concrete edges, wild grasses, ferns, gnarly small tree, the warmly glowing amber sphere, and the middle-aged man in a loose brown suit playing a brass saxophone.
Cinematic, highly naturalistic documentary scene captured on 16mm film, featuring fine grain, soft photographic detail, extremely low contrast, muted natural colors, lifted shadows, gentle highlights, and soft, diffused overcast daylight. Absolutely no HDR, no dramatic contrast, no glossy digital rendering, and no artificial movie lighting.
The platform floats in the center of a vast, calm lake surrounded in the distance by mist-shrouded forests. A thin layer of fog hovers over the water’s surface. The sky is an overcast grey-blue, establishing a muted, near-monochromatic atmosphere. The warm amber sphere casts subtle contrast, gently illuminating nearby foliage and reflecting off the lake’s surface.
The opening shot starts closer to the saxophone player than in the reference image, while keeping the platform, glowing orb, trees, and surrounding lake in view. The man continues playing the saxophone naturally, accompanied by subtle breath movements, finger shifts, and gentle swaying of the body.
The camera operates independently, free to explore the entire environment, gliding smoothly and controlled across the water. The camera is not tethered to the character or platform. Maintain authentic spatial continuity and natural camera motion throughout.
Across the entire experience, all elements must strictly maintain photorealism, physical consistency, documentary-grade aesthetics, ultra-low contrast, and gentle desaturation.
Single continuous take. Realistic ambient lake sounds, distant wind, gentle water ripples, and authentic saxophone acoustics. No non-diegetic background music beyond the actual instrument present in the scene.
Prompt: Use the reference image as the exact first frame and visual baseline for the scene. Retain the identical floating rectangular platform, weathered concrete edges, wild grasses, ferns, gnarly small tree, the warmly glowing amber sphere, and the middle-aged man in a loose brown suit playing a brass saxophone.
Cinematic, highly naturalistic documentary scene captured on 16mm film, featuring fine grain, soft photographic detail, extremely low contrast, muted natural colors, lifted shadows, gentle highlights, and soft, diffused overcast daylight. Absolutely no HDR, no dramatic contrast, no glossy digital rendering, and no artificial movie lighting.