We're on a mission to make 3D AI as accessible as 2D image generation.
After 13 years in AI research, we've watched AI master text (ChatGPT), images (Midjourney), and video (Sora) — but not 3D.
Apple's Vision Pro and Meta Quest made it clear: spatial computing will be next.
We're building the foundational software and APIs for spatial intelligence — enabling developers to reconstruct, edit, and generate navigable 3D worlds as easily as they work with 2D images today.
Apple and Meta have validated spatial computing as a market with Vision Pro and Quest 3. The hardware is here — now we need the software infrastructure.
Gaussian Splatting (2023) made real-time 3D reconstruction feasible. Neural rendering and diffusion models can now extend to spatial generation.
Generative AI technology is mature enough to extend beyond 2D. Transformers, diffusion models, and neural rendering converge for 3D.
Deep expertise in AI research, computer vision, and spatial intelligence
Founder & CEO
Associate Professor at UC Davis, 13 years in AI research with 200+ papers in computer vision and multimodal AI. Founded the IFM Lab focused on multimodal intelligence. Leading product strategy, fundraising, and business operations.
PhD Researchers from UC Davis
Our founding engineers are PhD candidates specializing in spatial AI, 3D reconstruction, gaussian splatting, and generative modeling. Deep domain expertise in computer vision and neural rendering.
Our proprietary VGGT (Visual Geometry Grounded Transformer) architecture represents breakthrough research in multimodal spatial understanding, trained on 50M+ spatial samples.
$3M+ in research grants from the National Science Foundation supporting foundational research in spatial AI.
GPU compute credits for model training on A100 and H100 clusters via NSF ACCESS allocation.
Academic research collaboration with UC Davis Computer Science department and IFM Lab.
Currently raising a $5M seed round (target close Q1 2026). Early conversations with Tier-1 VCs and strategic angels from Meta/Apple.
Exabyte Spatial AI founded with core team from UC Davis. Initial prototypes for text/image-to-3D generation.
Spatial AI API alpha with 100 beta developers. Core reconstruction and generation endpoints available.
Public API beta with commercial pricing. Exabyte Studio software alpha release for early customers.
Exabyte Studio public beta. Complete platform with APIs, SDKs, and professional tools.
Interested in custom deployments, white-label solutions, or volume licensing?
Looking to partner with us? We're interested in hardware, platform, and strategic partnerships.
Journalists and media inquiries. Press kit and company assets available upon request.
Currently based in Davis, California (UC Davis).
Moving to San Francisco Bay Area after fundraise for proximity to tech ecosystem, investors, and customers.
January 2026
We're excited to announce the alpha launch of our Spatial AI API. 100 beta developers will get early access to reconstruction and generation endpoints.
November 2024
Exabyte Spatial AI officially founded by Professor Jiawei Zhang and founding team from UC Davis. Building the foundational platform for spatial intelligence.
We're hiring exceptional engineers, researchers, and designers to build the future of spatial intelligence