Genie 4
World model
Price not published
Genie 3
DeepMind’s general-purpose world model for interactive environments and agent training. Not callable via API; try it at labs.google/projectgenie (AI Ultra).
PricePrice not published
Where it stands
5newer world models since its release
- Genie 427 Oct 2026
- Xiaomi-Robotics-U0 (38B) / U0-4B13 Jul 2026
- Cosmos 3 (Nano / Super)1 Jun 2026
- 1X World Model (1XWM)12 Jan 2026
- Cosmos Predict 2.5 / Transfer 2.56 Oct 2025
- After it from Google
- Genie 427 Oct 2026
- Weights
- Closed weights.
How to call it
Where to use Genie 3.
| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Project Genie (Google Labs) | deepmind.google labs.google |
Pricing
Price not published.
No price source linked.
Specs
- Input
- Text and images
- Output
- Video
- Context window
- Not published
- Max output
- Not published
- Open weights
- No
- Licence
- Proprietary
- Released
- August 2025
- Status
- Preview
- Type
- World model
What stands out
What its maker marketed at launch, and what people found later, each with its source.
Real-time interactive world generation
First
Generates navigable, photorealistic 720p worlds at 20-24 fps from text/image prompts.
World memory / consistency
Regions stay consistent when revisited; multi-minute visual consistency.
Promptable world events
Change weather or introduce objects/characters mid-exploration via text.
Consumer world sketching and remixing
Found after launch
Project Genie (2026-01-29) lets users sketch, explore and remix worlds (60 s sessions), combining Genie 3 with Nano Banana Pro and Gemini.
Notes
No public API. Research preview announced Aug 2025; consumer access via Project Genie only for Google AI Ultra subscribers, US, 18+ (not Business accounts).
Sources: Genie 3 page, Project Genie blog.
Not yet checked against an official source: Exact announcement day not re-verified.