What is Project Genie?
Project Genie is Google Labs' experimental research prototype for creating and remixing interactive worlds, powered by Google DeepMind's Genie 4. Official access at launch is for eligible Google AI Ultra subscribers 18 and over, with a waitlist for Google AI Pro. The official product page checked on 21 November 2026 does not list a dollar price.
What Project Genie is for
Project Genie is a Google Labs prototype for creating, exploring, and remixing interactive worlds. The official product page, checked on 21 November 2026, says it is powered by Google DeepMind's latest world model, Genie 4, and that you can shape worlds in real time, build with a Project Genie agent, and explore or remix a community gallery. Google describes it as an experimental research prototype.
The official 18 November announcement says Genie 4 lets users prompt mid-session to change a world, such as shifting the weather or adding a character, and that it pairs those visuals with real-time native audio, from ambient soundscapes to character dialogue, for the first time. A Project Genie agent takes a natural-language description and, Google says, orchestrates multimodal models to create initial visuals and custom background scores. All video and audio generated in Project Genie includes SynthID watermarking.
Genie 3 remains in Project Genie's model picker for longer runs or panorama-seeded spaces. Official Genie 3 updates include sessions of up to 3 minutes per generation, linking multiple worlds, 360-degree panorama image grounding, and real-time in-world events. The 3-minute figure is stated for Genie 3, not for Genie 4.
A first path to try it
If you are an eligible Google AI Ultra subscriber aged 18 or over, the official starting point is labs.google/projectgenie, which offers an Explore now path. If you are a Google AI Pro member, the same page offers a waitlist. The official pages do not list a country-by-country eligibility table, so this profile does not invent one.
On 19 November 2026, Google also said Genie 4 is available for developers through the Gemini API, via the Interactions API, and in Google AI Studio. That post says you can start from a text prompt, a reference image, or both, and experiment with continuous 720p worlds at 24 frames per second with synchronized native audio, plus real-time keyboard controls and live text prompts. It points readers to pricing details without stating a dollar amount in the post.
This is a suggested official path, not a recorded test. After a first session, check which model the picker used (Genie 4 or Genie 3), whether SynthID is documented for that output, and whether you were on the plan you intended.
Pricing, access, and limits
The official product page states that Project Genie is available at launch for eligible Google AI Ultra subscribers, that Google AI Pro members can join the waitlist, and that the prototype is rolling out to Ultra subscribers aged 18 and over. Those pages, checked on 21 November 2026, do not list a dollar price for Ultra. A linked plans page was not used as a price source in this session, so no subscription amount is quoted.
The official Gemini API announcement from 19 November 2026 tells developers to review pricing details, but the post itself does not include a token or session price. This profile therefore does not invent an API rate. Search interest in a free Project Genie plan is not supported by the official pages checked today, which describe Ultra eligibility and a Pro waitlist rather than a free tier.
Google still describes Genie 3 and Genie 4 as world models in frontier research, with work ongoing on camera latency, physical interactions, real-time audio precision, and long-term scene permanence. Inputs and outputs go through safety screenings, and official outputs include SynthID watermarking.
What stands out
- Official access at launch for eligible Google AI Ultra subscribers, with a Google AI Pro waitlist, as stated on labs.google/projectgenie
- Genie 4 mid-session prompting to change a world, plus real-time native audio, as described in the 18 November official announcement
- A Project Genie agent that, Google says, turns a natural-language description into initial visuals and custom background scores, plus a community gallery to explore and remix worlds
- Genie 4 is also officially available to developers through the Gemini API via the Interactions API and in Google AI Studio, as of 19 November 2026, with SynthID watermarking on outputs
Before you choose
- Official pages checked on 21 November 2026 do not list a dollar price for Google AI Ultra or for Genie 4 API use, and they do not define which countries or accounts are eligible
- The official 3-minute session length is stated for Genie 3, not for Genie 4
- Google still describes the product as an experimental research prototype, with work ongoing on camera latency, physical interactions, real-time audio precision, and long-term scene permanence
- This profile is not a hands-on test of Project Genie or the Gemini API
Go to the source
Sources & further reading
Based on the official documentation below. Features and access can change; check the provider before making a decision.
- Official Project Genie product page on labs.google, rendered 21 November 2026 (opens in a new tab)
- Google: Genie 4 is now available, dated 18 November 2026, rendered 21 November 2026 (opens in a new tab)
- Google: Genie 4 now available through the Gemini API, dated 19 November 2026, rendered 21 November 2026 (opens in a new tab)
- Official Google DeepMind Genie 4 page, rendered 21 November 2026 (opens in a new tab)
Was this page useful?
Help us make our coverage more useful. This is feedback on the page, not a rating of the tool.
Go a little further