Odyssey has raised USD 310 million in Series B funding to accelerate the development of general world models, a new class of artificial intelligence systems designed to understand, predict, and simulate the physical world.
The funding round values the company at USD 1.45 billion and positions Odyssey among the most closely watched AI labs working beyond traditional large language models. The round was led by Natural Capital, with participation from Amazon, AMD Ventures, GV, EQT, and In-Q-Tel. The company also attracted prominent individual investors including Jeff Dean, Elad Gil, Qasar Younis, Garry Tan, Guillermo Rauch, and Kyle Vogt.
Odyssey is building what it calls general world models. Unlike language models that predict the next word in a sentence, world models are designed to predict how physical environments behave. They learn how objects move, how light changes, how spaces respond to action, and what happens when something is touched, moved, or pushed.
This approach marks a major shift in artificial intelligence research. For several years, large language models have dominated the AI industry through chatbots, writing tools, coding assistants, and productivity applications. Odyssey’s work reflects a growing belief that the next frontier of AI may depend on systems that can understand and simulate the real world, not just generate text.
World models could have wide applications across robotics, science, healthcare, education, gaming, defense, and industrial systems. A robot could use a simulated environment to rehearse a task before performing it in the real world. Researchers could test ideas inside a model of a physical process before conducting experiments. Developers could build interactive simulated environments where humans and AI agents operate together.
The new funding will be used to scale Odyssey’s compute power and push its research closer to real-world deployment. Alongside the financing announcement, Odyssey also entered into a partnership with Amazon Web Services, which will become the company’s preferred cloud provider.
Through the AWS partnership, Odyssey plans to use AWS Trainium chips, which are built for advanced AI workloads. World models require significant computing resources because they must process complex physical interactions, long-term prediction, multimodal data, and real-time simulation. The additional compute capacity is expected to help Odyssey train more capable systems and expand its research.
Amazon’s participation in the round also highlights the strategic importance of infrastructure in the next phase of AI development. As models become more complex, access to specialized chips, cloud systems, and high-performance computing environments will be critical for companies working at the frontier of AI research.
Odyssey was founded by veterans of the self-driving car industry, a background that strongly shapes the company’s approach. Self-driving technology depends on the ability to interpret physical environments, predict movement, understand risk, and make decisions in real time. Odyssey is extending that idea beyond vehicles by working toward general systems that can simulate the physical world itself.
The company is led by Co-Founder and CEO Oliver Cameron. Its research team includes talent from leading technology and AI organizations such as DeepMind, Tesla, Waymo, Meta, Apple, and Wayve. Members of the team have worked on major AI and autonomy projects, including Gemini, Veo, GAIA world model research, and Tesla’s self-driving software.
Odyssey has already released several research projects that show different parts of its broader vision. Odyssey-2 Max focused on improving the physical accuracy of world simulations. Starchild-1 introduced real-time multimodal interaction. Agora-1 allowed multiple users or AI agents to share the same simulated environment. PROWL explored how a world model can improve itself through active exploration rather than passive observation.
These projects show that Odyssey is not only focused on building a single AI product. Instead, it is developing a broader research foundation for world simulation as a general AI capability. The company believes world models could become a new form of foundation model, similar in importance to large language models but designed for physical understanding and interaction.
Natural Capital said its investment was based on confidence in Odyssey’s research direction, technical leadership, and execution. The firm described the funding as its largest investment to date, reflecting strong investor belief that world models could help define AI beyond language-based systems.
The participation of Amazon and AMD Ventures also suggests that world model development is becoming strategically important for companies involved in AI infrastructure and advanced computing hardware. These systems require high-performance chips and scalable cloud resources, making partnerships with major compute providers increasingly valuable.
The size of Odyssey’s Series B round reflects growing investor interest in AI models that can move beyond text generation. While large language models remain central to the AI market, investors are now looking toward technologies that can support robotics, simulation, automation, physical reasoning, and real-world decision-making.
Odyssey’s goal is to help world model research reach what its CEO described as a “GPT-3 moment.” This refers to the kind of breakthrough that transformed large language models from a research area into a widely recognized technology category. If world models reach a similar stage, they could become central to the future of robotics, synthetic environments, scientific discovery, and intelligent systems that interact with the physical world.
The company’s USD 310 million funding round marks an important step toward that goal. With major investor backing, strategic cloud support from AWS, and a research team rooted in autonomy and advanced AI, Odyssey is preparing to scale its work in one of the most technically demanding areas of artificial intelligence.
As AI continues to evolve, the industry is moving from systems that understand language toward systems that can reason about space, motion, causality, and physical interaction. Odyssey’s latest funding shows that world simulation is becoming one of the most important emerging frontiers in artificial intelligence.
The round signals strong confidence in the future of general world models and their potential to support real-world applications across multiple industries. If successful, Odyssey’s technology could help shape the next generation of AI systems capable of understanding not only what people say, but how the world itself works.
