Odyssey raises $310 million at $1.45 billion valuation to build AI that simulates the real world

Odyssey, a startup building what researchers call “world models,” announced Wednesday that it has raised $310 million in Series B funding at a $1.45 billion valuation. The round was led by Natural Capital and included participation from Amazon, AMD Ventures, GV, EQT, In-Q-Tel, and other investors.

For the past few years, the AI industry has been obsessed with language models. A growing number of researchers now believe the next frontier is world models, AI systems that learn to simulate and predict the physical world the way large language models predict text.

The company is betting on a different path from the one taken by most of today’s AI leaders. Instead of predicting the next word in a sentence, Odyssey’s models are trained to predict what happens next in the physical world.

That idea traces back to the company’s roots in autonomous driving.

Odyssey was founded by CEO Oliver Cameron and CTO Jeff Hawke, who previously worked on self-driving technology, one of the toughest prediction problems in AI. A self-driving system must constantly anticipate how vehicles, pedestrians, cyclists, weather conditions, and road environments will change from one moment to the next.

The founders came away with a larger idea. If AI can learn cause and effect on the road, perhaps it can learn it everywhere.

That vision has attracted some of Silicon Valley’s most influential investors and AI leaders. Existing backers include Jeff Dean, Google’s chief scientist; investor Elad Gil; Applied Intuition CEO Qasar Younis; Y Combinator CEO Garry Tan; Vercel founder Guillermo Rauch; and Cruise founder Kyle Vogt.

Forget Text Generation: AI Startup Odyssey Raises $310M to Teach AI How the Real World Works

The funding arrives as interest in world models gains momentum across robotics, gaming, scientific research, and autonomous systems. Researchers increasingly see simulation and prediction as key ingredients for building AI systems that can interact with the physical world rather than simply generate text, images, or code.

Alongside the funding announcement, Odyssey said it has entered into a strategic agreement with Amazon Web Services, which will become the startup’s preferred cloud provider.

The partnership gives Odyssey access to AWS infrastructure, including Trainium chips, Amazon’s AI accelerators built for large-scale machine learning workloads.

“We believe world models represent a new class of foundation model—AI that can understand and simulate the world itself,” said Oliver Cameron, Co-Founder and CEO of Odyssey. “The last few years have seen major breakthroughs in scaling, interactivity, multimodality, and physics accuracy, and the field is now advancing extremely quickly. This round provides the compute, infrastructure, and partners to push the frontier of general world models, and to achieve a GPT-3 moment for the field.”

Amazon sees world simulation as one of AI’s most demanding workloads.

“World models represent one of the most demanding workloads in AI—they require massive compute throughput with tight latency constraints,” said Ron Diamant, Vice President and Distinguished Engineer at Amazon. “Odyssey’s team has been pushing the boundaries of what’s possible in this space, and Trainium is purpose-built for exactly this kind of scale. We’re excited to support this next phase of growth with AWS as Odyssey’s preferred cloud provider, collaborate on optimizing their models on our silicon, and work together to help accelerate applications in robotics, gaming, science, and beyond.”

Over the past three years, Odyssey has steadily published research aimed at moving world models from theory into practical systems. The company says Odyssey-2 Max improved physics accuracy in general world simulation, Starchild-1 introduced real-time multimodal world models, and Agora-1 enabled multiple humans and AI agents to interact inside a shared simulated environment where actions persist and affect everyone involved.

Those efforts point to a future in which AI agents can learn, test ideas, collaborate, and make decisions within virtual worlds before acting in the real world.

Investors appear convinced that the future could become one of the next major chapters in AI.

“At Natural Capital, we invest behind ambitious technical teams building what comes next,” said Jay Zaveri, General Partner at Natural Capital. “We developed deep conviction in Odyssey’s research direction, technical leadership, and execution, which made this our largest investment to date. We believe they have the potential to help define AI beyond language models.”

For now, large language models remain the dominant force in AI. Yet a growing number of researchers believe language alone may not be enough to reach more capable forms of intelligence.

That belief is fueling a surge of interest in world models, systems that learn physics, causality, spatial reasoning, and long-term prediction. Odyssey’s latest funding round suggests investors are willing to place a sizable bet on that vision.