Odyssey raised $310M to build worlds, and it picked AMD over Nvidia
World-model startup Odyssey closed a $310M Series B at a $1.45B valuation, led by Natural Capital with Amazon, AMD Ventures, and GV in the round. The founders are self-driving veterans, CEO Oliver Cameron built Voyage before it went to Cruise, CTO Jeff Hawke came out of Wayve. They're not building another chatbot. World models gather data from the physical world and simulate it with real physics, the bet that the next layer of AI is interactive worlds, not more text.
The detail I keep circling is the silicon. Odyssey took Nvidia's money in an earlier round and then turned around and made AWS its preferred cloud, leaning on Amazon's Trainium chips. A frontier lab publicly choosing the challenger stack over the default one is a small thing that says a lot. When even well-funded labs start routing around Nvidia, the compute monopoly story gets its first real cracks.
This is the funding round, not the product. Odyssey shipped Agora-1 earlier, this $310M is the fuel to scale the simulation work. World models are still the riskiest corner of the frontier, hard to evaluate, hard to monetize today, easy to over-hype. But Amazon and AMD writing checks into one means the physical-AI thesis has serious institutional backing now, not just demos.
Here's the bigger frame. Text models are converging, everyone's within a few benchmark points of each other. The labs chasing a real moat are the ones leaving language behind, into worlds, into bodies, into simulation. Odyssey just got $310M to find out if that's where the next 100X lives. Details at https://techcrunch.com/2026/06/17/world-model-maker-odyssey-nabs-1-45b-valuation-backed-by-amazon-and-other-big-names/.
← Back to all articles
The detail I keep circling is the silicon. Odyssey took Nvidia's money in an earlier round and then turned around and made AWS its preferred cloud, leaning on Amazon's Trainium chips. A frontier lab publicly choosing the challenger stack over the default one is a small thing that says a lot. When even well-funded labs start routing around Nvidia, the compute monopoly story gets its first real cracks.
This is the funding round, not the product. Odyssey shipped Agora-1 earlier, this $310M is the fuel to scale the simulation work. World models are still the riskiest corner of the frontier, hard to evaluate, hard to monetize today, easy to over-hype. But Amazon and AMD writing checks into one means the physical-AI thesis has serious institutional backing now, not just demos.
Here's the bigger frame. Text models are converging, everyone's within a few benchmark points of each other. The labs chasing a real moat are the ones leaving language behind, into worlds, into bodies, into simulation. Odyssey just got $310M to find out if that's where the next 100X lives. Details at https://techcrunch.com/2026/06/17/world-model-maker-odyssey-nabs-1-45b-valuation-backed-by-amazon-and-other-big-names/.
Comments