genie 4
Describe a scene and Genie 4 builds an explorable, photoreal world where physics, sound and character dialogue react to you live.
Report an Issue
AI Dance Tools
Explore our AI-powered dance video creation tools
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
The Future of AI Video Is Here.

Seedance 2.0
The Future of AI Video Is Here.

Seedance 2.0
The Future of AI Video Is Here.

Gemini Omni
Gemini Omni Video Generator
Image to Video
Generate videos from images

Gemini Omni 1.2 Flash

AI Multi-Scene Shorts Generator
Create viral AI Shorts instantly
Meet the Engine Powering Genie 4
Give it a sentence or a single still frame, and Genie 4 returns a walkable environment in which image, motion, sound and spoken lines are produced as one.
- One Model, Every ModalityAmbience and voices are produced together with the visuals from the start, rather than layered on after the environment already exists.
- Steer It While It RunsChange the weather, swing the camera or bring in another character mid-session; Gemini adapts the result to whatever is currently on screen.
- Gemini at the CoreGemini Omni anchors appearance and motion in reality, while companion Gemini models track consistency across successive frames.
How to Get Started with Genie 4
Four short moves take you from a waitlist request to a scene you can steer as it plays.
What Genie 4 Can Do
The strengths a modern world simulator brings to live, interactive scenes.
Convincing Physics and Collisions
Objects carry plausible weight and momentum, with airflow, impacts and contact behavior all handled inside the simulation.
Lifelike Human Motion
Figures walk, gesture and talk in a natural way, and you can redirect them on the spot without pausing the scene.
Sound Built Into the Scene
Every location ships with its own soundtrack — wind through branches, waves hitting the shore — generated alongside the imagery.
Conversations That Respond
Characters reply to what you type and to events unfolding around them, and their speech keeps pace as the situation develops.
Planning Across Many Steps
Gemini 3.8 Flash sits in the control loop, mapping out actions, reviewing rendered frames and revising the plan as conditions change.
Cinematic Photoreal Look
Film grain, lighting and lens behavior remain steady, so long takes and free camera movement still look like real footage.
Genie 4: Your Questions Answered
Straight answers about how the Genie 4 world model works and who can use it.
So what is Genie 4, in plain terms?
It is a multimodal world model: instead of just drawing a frame, it predicts how a scene will move, sound and react a moment later.
Where can I try it?
Access is limited to a research preview for vetted testers. Sign up for the Project Genie waitlist, or watch for openings via the Gemini API and AI Studio.
Will it produce audio and spoken lines?
It does. Background sound and character dialogue are generated by the model itself and shift in response to what you ask for.
How do I direct what happens?
Keyboard input lets you roam the environment, and typed prompts let you nudge the story forward while the simulation continues running.
Why does it rely on Gemini?
Gemini Omni keeps appearance and motion realistic, and additional Gemini models maintain continuity so the world does not drift apart over time.
What kinds of projects suit it?
Robotics and agent training grounds, navigation benchmarks for embodied AI, and interactive entertainment that reacts to the player.
Request Early Access to Genie 4
Add your name to the Project Genie waitlist for Genie 4, or look into early access through the Gemini API and Google AI Studio.
