How to Access Genie 4 Interactive World Models
Step-by-step guide to Genie 4 features, real-time steering, native audio, and joining the Project Genie waitlist for early access.
Google DeepMind announced Genie 4 on October 27, 2026. The model generates interactive worlds you can steer in real time with text prompts while it produces matching audio, including ambient sound and dialogue.
It builds on earlier Genie models and incorporates Gemini capabilities for coherence. Access remains limited to a research preview for trusted testers, with a public waitlist open for Project Genie. Google AI Ultra subscribers and selected Gemini API partners are next in line.
This guide covers what Genie 4 does, the current access path, and practical steps to join the waitlist based on official announcements.
What Genie 4 adds
Genie 4 is a real-time interactive world model. You describe a starting scene or upload a reference image, then use keyboard controls to move and text prompts to change the environment while it runs. Gemini models help maintain continuity so objects and characters stay consistent as the scene evolves.
Key improvements over Genie 3 include:
- Native audio generated alongside the visuals, covering ambient sounds, music, and interactive dialogue.
- Real-time steering with natural language. Change the weather, move the camera, or introduce a new character without restarting.
- Longer coherent scenes, up to roughly two minutes in demos versus one minute previously.
- Stronger photorealism and more natural human movement and reactions.
Official demos show a saxophone player on a floating platform over a lake, a flowing pink fabric river through a meadow, and other scenes that respond to mid-run prompts. All audio in the demos comes directly from the model.
How access works today
Genie 4 is not open to the public. Google DeepMind is rolling it out in stages:
- Trusted testers across filmmaking, game design, education, journalism, and robotics receive it first for feedback.
- Google AI Ultra subscribers gain access next inside Project Genie, the existing experimental web app that previously ran Genie 3.
- Early access partners can use it through the Gemini API and Google AI Studio.
No public release date, pricing, or model identifier has been published. Project Genie currently requires a Google AI Ultra subscription and is limited to users 18 and older in supported countries. Genie 4 itself is described as coming soon to Ultra subscribers in that app.
Anyone can join the waitlist now. The form is linked from the DeepMind Genie 4 announcement page.
Step-by-step: join the waitlist
Follow these steps to request access:
- Open the official Google DeepMind Genie 4 announcement at deepmind.google/blog/genie-4-a-new-generation-of-world-simulation/.
- Click the “Join the Genie 4 waitlist” button near the top of the page.
- Sign in with your Google account when the form loads. The form title is typically “Apply for Genie 4 access in Project Genie.”
- Complete the required fields and submit. Google has not published selection criteria or expected wait times.
- Monitor your email for any follow-up. If you already hold a Google AI Ultra plan, also check Project Genie periodically once the rollout begins.
The waitlist form is free to submit. Access to the full experience, when granted, is expected to remain tied to the Ultra plan for consumer users.
What to expect once you have access
In Project Genie you describe an environment and character, then explore the generated world. Keyboard controls handle movement and camera; text prompts steer the scene in real time. Genie 4 adds the audio layer so the world produces matching sounds as it runs.
Gemini sits in the loop to interpret your prompts, track the world state, and plan multi-step changes so the simulation stays coherent. Early use cases highlighted by DeepMind include VFX pre-visualization, game prototyping, educational scenarios, and safe robot training environments.
Practical tips while you wait
- Join the waitlist early. Selection is not guaranteed and no timeline is public.
- Review official Project Genie help documentation so you understand current limits (age 18+, Ultra required, session constraints on the Genie 3 version).
- Watch the official DeepMind demo video to see keyboard controls and prompt steering in action before you receive access.
- Keep prompts specific when testing similar tools. Include visual style, lighting, and exact actions you want the scene to perform.
- For related Gemini document workflows, see our guide on how to generate, preview, and edit files in Gemini.
Limitations to know
Longer simulations can still drift in object permanence and fine details. Genie 4 improves coherence but remains a research preview. No public API pricing or self-serve endpoint exists yet. Consumer access routes through Project Genie and the Ultra plan; there is no free unlimited tier announced.
Always treat generated content as experimental. Review outputs carefully before using them in professional or public work.
Conclusion
Genie 4 advances interactive world simulation with real-time text steering and native audio. Access is currently limited to trusted testers, with a free waitlist open for Project Genie and Ultra subscribers next in line. Join the waitlist from the official DeepMind page, review the demo video, and prepare specific prompts so you can test the controls effectively when access arrives.
Official details and the waitlist link are on the Google DeepMind Genie 4 blog post. Check Google AI plan pages for current Ultra requirements and Project Genie availability in your region.
Sources
- Google DeepMind: Genie 4: A new generation of world simulation (October 27, 2026)
- Google Blog: We’re announcing Genie 4 (October 27, 2026)
- Official YouTube demo from Google DeepMind
- Project Genie support documentation on Google Labs
Share this story
