Matthias Niessner, one of Europe’s top AI researchers and cofounder of avatar startup Synthesia, has secured a $13 million seed round for his new venture, SpAItial. Headquartered in Munich, the company is building foundation models that transform written prompts into highly realistic and interactive 3D environments. The round was led by Earlybird Venture Capital, with Speedinvest and a group of notable angel investors also participating. It stands among Europe’s largest seed rounds for an AI startup. Though SpAItial AI hasn’t launched a product yet, an early teaser, a detailed 3D-rendered room created from a simple sentence, hints at its big ambitions. The founding team blends deep academic research with practical industry expertise and aims to lead Europe’s next wave in generative AI.
SpAItial AI’s Engine Promises More Than Just Realism
SpAItial AI is building a model that goes beyond visuals. It’s not just about creating lifelike 3D scenes, it’s about making them behave like real environments. A user could type, “a sunlit kitchen with glass cups on the counter,” and get a scene with proper lighting, textures, and physics. That scene could be used in simulations, games, or animations.
“We don’t just want a world that looks real,” said Niessner. “We want it to behave like the real world.” He’s joined by Ricardo Martin-Brualla, who worked on Google’s 3D telepresence project Beam, and David Novotny, who led Meta’s text-to-3D efforts. Together, they bring years of experience in geometry, rendering, and machine learning. Their goal: make 3D creation as simple as typing a sentence.
From Vision to Use Case, with Big Potential and Bigger Unknowns
SpAItial AI plans to license its API to developers and creative studios. The tech could be applied in gaming, architecture, film, and robotics. Luke Rogers, a former Cazoo executive and longtime friend of Niessner’s from Stanford, has joined as cofounder to lead business operations.
Monetizing the vision will be tricky. Early users may need to work with rough outputs, while others may wait for more polished results. Competing ventures like Odyssey and World Labs, founded by AI luminary Fei-Fei Li, are exploring similar spaces. Still, SpAItial believes it has an edge. Competitors often focus on static imagery or minimal interactivity. SpAItial is aiming for full-scale, explorable 3D worlds.
Typing a Prompt Could Soon Build an Entire Game
The dream is bold: let a 10-year-old build a video game in 10 minutes using only text. Niessner thinks this is achievable, especially as platforms like Roblox still limit third-party creativity. If SpAItial succeeds, it could reduce reliance on complex tools like CAD software. The barrier to entry for interactive content creation would drop dramatically. For now, the focus remains technical. The team is working on object physics, lighting realism, and full-scene interactivity. Hiring is happening slowly and strategically. They want a small, fast-moving group that can focus on building foundational tech.
Is This the Start of a New Creative Era for 3D AI?
If SpAItial AI achieves its vision, it won’t just speed up 3D content creation. It could fundamentally change who gets to create in the first place. But that shift also comes with responsibility. Regulators and developers will need to define rules around ownership, ethical use, and platform integration. SpAItial isn’t focused on policy yet. For now, it’s building. But the path forward will demand thoughtful decisions. The opportunity is massive, and the race is clearly on.
About The Author
