Startups & Funding
Odyssey raises $310M to build world model AI
Odyssey, a U.K.-based world model AI startup, raised $310 million at a $1.45 billion valuation, and will optimize its models for AWS Trainium chips.
Odyssey, an artificial intelligence startup based in the U.K., has raised $310 million in a Series B funding round—the second major round of venture capital financing for a startup. The investment round was led by venture capital firm Natural Capital and values the company at $1.45 billion. Odyssey is focused on developing “world models,” which are AI models that simulate physical world physics. These models represent a development in artificial intelligence beyond text- and chat-based large language models.
The startup was founded in 2023 by CEO Oliver Cameron and CTO Jeff Hawke. Cameron’s previous startup, Voyage, was acquired by Cruise, while Hawke was formerly employed by the U.K.-based Wayve. This background in technology development supports the company’s current focus on world models. Odyssey offers its world models for a variety of use cases, including video-game creation and robotics. The company is perhaps best known for producing rich, interactive video from text prompts.
As part of the funding arrangement, Odyssey has named AWS as its preferred cloud provider. The startup will optimize its models to run on AWS Trainium chips, which are AWS-designed machine learning chips that compete with Nvidia’s AI chips. Amazon participated in the funding round as an investor and cloud provider, alongside other investors including AMD Ventures and GV, while Natural Capital served as the lead investor.
With this latest round, Odyssey has raised a total of $337 million to date. In addition to institutional venture capital firms, the startup secured backing from a group of individual angel investors, including:
- Jeff Dean
- Elad Gil
- Garry Tan
- Guillermo Rauch
- Kyle Vogt
Why it matters
Odyssey’s funding highlights the shift toward world models—AI that simulates physical physics—as the next frontier beyond text-based large language models, with major cloud providers like Amazon betting on the infrastructure required to run them.