The first thing you notice is the pause. You're watching a video of a dancer, a standard clip on a webpage. You hit the spacebar. The motion freezes, but the scene doesn't. You drag the mouse left, and the entire perspective swings, revealing the curve of her spine from a side angle you were never meant to see. This is the core promise of 4DV.ai: turning flat footage into a space you can inhabit, all without leaving your browser [4DV.ai].
Founded in 2024 by Jiaming Sun, a researcher from Zhejiang University’s influential 3D Vision Group, 4DV.ai is building a full-stack toolkit for what it calls 4D video. The technology, 4D Gaussian Splatting, reconstructs moving scenes from multi-camera footage into a cloud of tiny, time-aware data points. The result is an asset that allows for six degrees of freedom, letting a viewer pan, tilt, zoom, and orbit a moment in time as if it were a physical object [ruoqijin.com]. The company’s recent seed funding from First Round Capital and Sequoia Capital signals a bet that this niche academic technique is ready to become a practical medium for storytellers [Crunchbase].
From Academic Paper to Production Pipeline
Jiaming Sun’s background is the project’s technical bedrock. He is a co-author on foundational papers in neural volumetric video, including EasyVolcap and 4K4D [Google Scholar]. His most recent work, FreeTimeGS presented at CVPR 2025, directly underpins 4DV.ai’s approach, claiming a ~2.4 dB quality improvement over prior methods while achieving blistering speeds up to 467 frames per second on high-end consumer hardware [LinkedIn].
The startup is methodically building out the production stack required to move from dazzling demo to reliable tool. Its partnership with camera maker OBSBOT is the most tangible step. At the NAB Show 2026, the companies demonstrated a synchronized array of approximately 60 OBSBOT Tail 2 PTZ cameras feeding live into 4DV.ai’s reconstruction pipeline [OBSBOT, April 20, 2026].
The Browser as the Ultimate Distribution Layer
4DV.ai’s most consumer-friendly bet is on the browser itself. The entire rendering engine runs on WebGL, using a lightweight framework called PlayCanvas. There is no app to download, no special headset required. A user simply clicks a link and gains the ability to manipulate the viewpoint of a video. The company claims the experience is fast and fluid, with integrated spatial audio that changes as the user’s perspective shifts [websiteaid.in].
- Accessible Creation. The platform can work with existing 2D footage, theoretically allowing creators to retroactively add volumetric depth to old content [websiteaid.in].
- Frictionless Playback. By running in any modern browser, it bypasses app store gatekeeping and compatibility headaches.
- Downstream Utility. The system can export depth data for use in dedicated VR and AR projects [websiteaid.in].
The Volumetric Video Competitive Field
| Company | Primary Approach | Key Differentiator |
|---|---|---|
| 4DV.ai | 4D Gaussian Splatting | Browser-based playback; full-stack from capture to interaction [4DV.ai]. |
| 4Dviews | Traditional volumetric capture | Industry pioneer with studio-based systems for high-end production [4Dviews]. |
| Arcturus | Volumetric video editing & streaming | Focus on post-production and compression for distribution [Crunchbase]. |
| Gracia AI | AI-generated 3D assets | Emphasis on synthetic creation rather than real-world capture [Crunchbase]. |
The Skeptic's Case and the Scaling Challenge
The risks here are not subtle. The technology, while impressive in demos, must prove it can handle the messy variability of professional production. The computational cost of processing feeds from 60 cameras is non-trivial, raising questions about the scalability and cost structure of the cloud pipeline required. Furthermore, the market for interactive video remains largely unproven. The company has yet to publicly name marquee media or enterprise customers, leaving its commercial traction an open question.
4DV.ai’s answer appears to be a focus on practicality and partnerships. The OBSBOT integration is a move to standardize and simplify capture. The next twelve months will be critical for transitioning from conference demo to paid workflow. Key milestones to watch include the announcement of launch partners in media or e-commerce, the detailing of its SaaS pricing model, and a likely Series A round to scale its engineering and go-to-market efforts.