Odyssey opens its Odyssey-3 world model to the public
The California startup's model generates interactive 3D-like worlds from text in real time and is pitched as a base for robotics, drones and games.
California-based AI company Odyssey has launched a public research preview of Odyssey-3, a generative “world model” that builds interactive environments from text prompts and lets users explore them in real time, The Decoder reported on 11 October.
What it does
- Simulates physical processes and predicts how a scene changes in response to actions.
- Is designed as a foundation for controlling systems — from robotic arms and drones in simulation to video-game characters.
- Developers can apply for API access.
Founders Oliver Cameron and Jeff Hawke first showed the model on 15 September; what is new is public access, more technical detail and benchmark results.
Benchmarks — with a caveat
On the WorldMark benchmark Odyssey-3 ranks first in three of four categories, including Third-Person Real (79.0), and third in First-Person Real (80.6). These figures come from the company’s own evaluation, not an independent test.
Also available in: Саха тылаРусскийEspañol中文Portuguêsالعربية