Hello, World
Introducing Odyssey, Hollywood-grade generative worlds
We love to tell stories
Humans have always been storytellers. Our earliest ancestors told tales around campfires. We created the novel, The Iliad, The Odyssey, The Vedas, The Analects, and One Thousand and One Nights. We broadcast pivotal moments to millions on television. We crafted cinematic masterpieces that shaped us, and pioneered visual effects that transported us to far-away lands. We created Casablanca, 2001: A Space Odyssey, The Godfather, Star Wars, Schindler’s List, Jurassic Park, Toy Story, Interstellar, and Dunkirk.
As technology has evolved, our stories have too. Yet, through it all, one thing has remained constant—great stories are the result of blood, sweat, and tears.
We need to decide how AI tells stories
We need to hold AI to higher standards. A short trip around the web will reveal that we’re inundated with low-quality AI-generated content. Content farms, spam bots, and even well-intentioned companies are using AI to churn out text and imagery, with the goal of gamifying algorithms and capturing your attention. If not done right, AI video generation could head in a similar direction, where we are inundated with random videos that have no spark or story. On a long-enough timespan, perhaps we become addicted to these junk-food videos, forgetting what high-quality human storytelling looks like. Perhaps humans are relegated to storywatchers—not storytellers.
At Odyssey, we reject this future. Humans telling stories is too important to our way of life, and professional storytellers have proven that they have so much to offer. Similarly, so does powerful AI, if we build it right. Instead of replacing humans with algorithms that optimize for clicks, we believe a new visual AI should be placed in the hands of professional storytellers. This visual AI should enable them to not only generate stunning video, but also to precisely direct it, and to tell the epic tales stuck in their head.
A breakthrough in visual AI is needed
To deliver this, we're pioneering Hollywood-grade generative worlds, where beautiful scenery, characters, lighting, and motion can be both generated and directed. This technology will deliver the game-changing efficiencies and creativity of AI, but importantly, give you full, fine-tuned control over every element in your scene. From stunning mountains, to lush vegetation, to beautiful rivers, to epic characters, all the way to the low-level materials, lighting, motion, and more. Simply put, it's a new way to create films and games.
This capability is made possible by going layers deeper than traditional text-to-visual models. Instead of training one model that restricts you to a single input and a single, non-editable output, we're training four powerful generative models that enable fine-tuned control over each major layer of visual storytelling. Specifically, models capable of generating high-quality geometry, photorealistic materials, stunning lighting, and controllable motion. Combined, these models enable you to rapidly generate scenes and shots exactly as you dreamt them. What’s more, we’re building workflows around these models tailored for power users, with first-class integrations into existing production workflows in use in Hollywood, gaming, and beyond.
We're early on this journey, but we're so excited about the potential of this technology. We’d like to invite you to join along, whether by , exploring a role with us, or by following us. We’re eager to meet storytellers and storytelling companies of all kinds, to learn from you, and to shape this technology together.
Generative worlds for film, gaming, and more
Build worlds, capture sequences
The best stories transport you to epic worlds. This is why world-building—real or fictional—is at the heart of our work at Odyssey. We’re training new generative models that enable you to generate cinematic worlds, with 3D control over scenery, characters, lighting, and motion. Once you’ve generated your world, you direct it, and then capture cinematic sequences within it.
Your scene, your way
Odyssey gives creatives 3D control over generated elements within your scene, whether geometry, materials, lighting, or motion. Scenes are comprised of promptable and manipulable layers that can operate independently, while also remaining context-aware. Or, if you'd prefer, simply prompt the scene itself.
Iterate until perfect
The world’s best storytellers experiment with hundreds of ideas until they find one they love. Then, they iterate relentlessly until the implementation is perfect. Odyssey’s generative workflows, paired with pro-grade version control, will enable you to work the same way.
Integrate with your workflows
We’re embracing standards that enable interoperability with existing, powerful workflows used in high-end productions. Everything you create is editable and exportable in standardized 3D file formats, like USD.
Built by a team of AI researchers, computer graphics experts, and Hollywood artists
Our team has delivered state-of-the-art AI and simulation systems at companies such as Cruise, Waymo, Wayve, Tesla, Microsoft, Meta, and NVIDIA. We also have deep experience in computer graphics research, with lead engineers from video games like Spore, SimCity, The Sims, Alien: Isolation, and the Tom Clancy series, alongside technical artists behind films like Dune 2, Godzilla, The Creator, Avengers: Age of Ultron, Alita: Battle Angel, and Jurassic World: Fallen Kingdom. Multiple members of our team have won BAFTAs.
We're backed by world-class investors
“I do feel like the fingerprints—the squishy fingerprints on the thing—are the thing that make it unbelievably cool. The best movies, my favorite movies, are where you can feel the hand of the filmmaker.”