Deals

The Money Moves: Where Capital Is Flowing in Spatial Computing

A wave of nine-figure rounds and a landmark acquisition have reshuffled the investor map for spatial computing. The categories drawing the most capital tell you where the industry thinks the next five years are being built.

Mainstay Communications Team · September 3, 2026 · 6 min read

In brief

Spatial computing attracted its largest private rounds on record between mid-2025 and mid-2026, with nine-figure checks flowing into generative 3D, AI-powered wearables, and contact-lens computing. The Niantic-Scopely deal restructured the field by freeing a dedicated geospatial AI company with $250 million in capital. Luma AI, Sesame, Even Realities, and XPANCEO pulled the largest rounds; Meta paid over $2 billion to acquire AI agent startup Manus. The capital is concentrating into fewer, larger bets on the infrastructure layers the category will run on.

The Niantic pivot that rewired the map

The deal that did the most to signal a new era didn't look like a spatial computing investment at all. In March 2025, mobile gaming firm Scopely agreed to buy Niantic's games business, including Pokémon GO, for $3.5 billion. The games were the asset. The point was what was left.

Niantic spun the remainder into a standalone company, Niantic Spatial Inc., capitalized at $250 million: $200 million from Niantic's own balance sheet, $50 million from Scopely. CEO John Hanke runs it. The focus is geospatial AI, a bet that the physical world needs to be mapped, indexed, and made machine-readable at scale.

The structure was a clean declaration of intent. Niantic's investors didn't need to be convinced that location-based gaming was a good business. What they decided was that the geospatial AI platform underneath the games was worth building separately, free from the quarterly pressure of running live titles.

Nine-figure rounds and what they're betting on

Three categories pulled the largest private rounds in the 12 months from mid-2025 through mid-2026: generative 3D, AI-native wearables, and the physical-world interface layer.

Generative 3D. In November 2025, Luma AI closed a $900 million Series C led by HUMAIN, the investment arm of Saudi Arabia's Public Investment Fund, with participation from AMD Ventures and existing backers including Andreessen Horowitz. The company's flagship platform, Dream Machine, generates professional-grade video and images. Luma's deal also included a partnership on Project Halo, a 2-gigawatt AI compute cluster in Saudi Arabia. That infrastructure detail matters: generating photorealistic 3D at scale is a compute-intensive problem, and the sovereign capital flowing into it suggests sovereign-level conviction.

Smart glasses. Two big rounds defined the wearables conversation. Sesame, founded by Oculus co-founder Brendan Iribe alongside former Ubiquity6 CTO Ankit Kumar, raised $250 million in a Series B led by Sequoia and Spark Capital in October 2025. Former Oculus co-founder Nate Mitchell joined the company as chief product officer. The product is AI-powered smart glasses with conversational voice interaction, with the core claim that on-device AI should feel like talking to a person rather than a command interface. Then in July 2026, Even Realities, a Shenzhen startup founded by ex-Apple engineers, raised $150 million in a pre-Series B led by Meituan and Tencent at a $1 billion valuation. Even's positioning is explicit: no cameras, privacy-first, display hardware aimed at professionals who won't wear a camera on their face all day.

Contact lens computing. The round that drew the widest coverage was XPANCEO's Series A in July 2025. The Dubai-based startup raised $250 million led by Opportunity Venture (Asia), reaching a $1.35 billion valuation. XPANCEO is building all-in-one smart contact lenses with XR, health monitoring, night vision, and wireless charging built into a lens. Road to VR called it the largest single AR round in history. The company had 15 working prototypes at the time of the raise. It's a long-term hardware bet, and the valuation price tags it accordingly.

The M&A layer

Acquisitions in the period were fewer but pointed in a clear direction: big platforms buying specialized AI capabilities they don't want to wait for.

Meta's $2 billion-plus acquisition of Manus in December 2025, reported by Bloomberg and confirmed by CNBC, is the clearest example. Manus is a Singapore-based AI agent startup that can execute complex research, coding, and data tasks autonomously. The purchase fits Meta's broader spatial computing strategy: the hardware (Quest headsets, Ray-Ban smart glasses) is already shipping; the AI layer that makes those devices useful is where the acquisitions are going.

Where the capital is concentrating

The broader market context sharpens the picture. According to Tracxn data, spatial computing companies raised $2.75 billion across 91 rounds in the first half of 2026, compared to $1.19 billion across 109 rounds in the same period of 2025. More money through fewer deals, which means larger checks and earlier filtering. In 2025 and into the first half of 2026, the top handful of deals captured the vast majority of the capital.

The practical consequence: there's almost no conventional $20-to-$50 million growth round happening in the sector right now. Companies are either small and seed-funded, or they're pulling in nine-figure checks from sovereign wealth funds, strategic corporate investors, and top-tier VCs. The middle is thin.

Three categories are drawing consistent investor interest below the mega-round level. Healthcare XR, where startups like ORamaVR and MediView XR are funding clinical and surgical training tools, has maintained steady deal flow because the enterprise ROI is measurable and the regulatory path, while hard, is understood. Enterprise workflow software, where companies like JigSpace are building spatial content tools for field service and sales teams, is attracting capital because it sells to buyers who already have a budget line for industrial training and field operations. And the content infrastructure layer, tools that make 3D content easier to create, convert, and deploy, is drawing attention from investors who watched the first wave of headset-first companies struggle and concluded that the bottleneck was authoring, not display.

What the pattern says

The concentration of capital tells the underlying story. Investors aren't spreading bets across the category. They're writing large checks into companies where the core question is technical feasibility rather than market existence.

XPANCEO and the contact lens bet is the longest-dated: a hardware wager that the form factor question resolves in its favor over the next 5 to 10 years. Luma AI and the generative 3D stack is a shorter-dated infrastructure bet: that the tools for creating spatial content will become as foundational as the cloud storage that holds it. Sesame and Even Realities are competing for the wearable interface that sits between your world and your AI.

What's not getting funded at scale is consumer VR content, standalone headset-native experiences, and general-purpose metaverse infrastructure. The market absorbed the lessons of the 2021 to 2023 cycle, and the investors who stayed in the category moved toward the layers with clearer revenue models and shorter timelines to enterprise contracts.

The money isn't guessing about spatial computing anymore. It's picking which infrastructure the category will eventually run on.

Sources

Companies mentioned

Mainstay Digital · Spatial and 3D visualization for industrial companies

Back to Newsroom