Odyssey raises $310 million Series B for world models at a $1.45 billion valuation

Odyssey, an AI lab building general-purpose world models, has raised $310 million in a Series B round at a $1.45 billion post-money valuation, the company said on June 17. The round was led by Natural Capital, with participation from Amazon, AMD Ventures, GV (Google Ventures), EQT, and In-Q-Tel, the venture fund affiliated with the U.S. intelligence community.

The valuation is a sign of how much investor money is flowing toward systems that aim to model the physical world rather than generate text or images. Odyssey builds AI that can understand and simulate physical environments, a problem set distinct from the language models that dominate current AI investment.

The deal also deepens Odyssey's ties to Amazon. The company said the agreement names AWS as its preferred cloud provider, and that it is working with Amazon's Annapurna Labs to optimize its models for AWS Trainium, Amazon's custom machine-learning accelerators. That pairing gives Amazon a foothold with a buzzy world-model lab while giving Odyssey a path to cheaper training and inference than general-purpose GPUs.

The investor mix is notable for who is in and who is out. AMD Ventures joins as a new shareholder. NVentures, NVIDIA's venture arm, had backed an earlier Odyssey round in February 2026 but is not part of this Series B group. Odyssey's founding team came out of autonomous-driving research, where it built systems that learned a model of the world including physics, causality, dynamics, and human behavior.

Odyssey did not disclose revenue, a product release timeline, or technical benchmarks for its world models, leaving the lab's progress largely unproven outside its own description. With $310 million in fresh capital and a dedicated chip-optimization effort, the company now faces the harder task of turning a well-funded research agenda into systems customers can use.