Meta’s metaverse bet has turned into one of the most expensive science projects in corporate history, with its virtual and mixed reality arm racking up losses on a scale that would sink smaller companies. Reality Labs, the division behind Meta’s headsets and VR software, lost roughly $19 billion last year and executives are warning investors that 2026 will look a lot like 2025. The company is trying to convince Wall Street that this cash bonfire is a necessary prelude to a new computing platform, even as it pivots its public story toward artificial intelligence and wearables.
The result is a strange split-screen moment. Meta’s core ad business is thriving and its AI investments are framed as a land grab, while its VR and metaverse push remains a deep, open-ended money pit. I see a company that is not backing away from the vision that prompted its name change, but is quietly redefining what that vision means and how long investors are expected to wait.
The $19 billion hole behind the metaverse dream
At the center of Meta’s extended reality strategy is Reality Labs, the unit responsible for Quest headsets, mixed reality software and the broader metaverse push. That unit’s $19.1 billion loss for 2025, compared with $17.7 billion in 2024, shows that the financial drag is not just persistent, it is getting worse. Another breakdown of the numbers puts Reality Labs’ full year revenue at $2.2 billion against a loss of $19.2 billion, underscoring how small the business remains relative to the resources poured into it. When a division generates roughly one dollar of sales for every nine dollars it burns, it is less a growth engine than a subsidized experiment.
Meta itself has acknowledged the scale of the problem. One analysis described Reality Labs as a $20 billion sinkhole, a phrase that captures both the magnitude of the losses and the uncertainty about when, or if, they will translate into durable profits. Another report on Meta’s metaverse division pegs the 2025 shortfall at $19.2 billion, reinforcing that, however you slice the accounting, the red ink is staggering. I read these figures less as a one-off misfire and more as evidence that Meta has locked itself into a long, expensive march toward a platform that consumers have not yet embraced at scale.
Layoffs, restructuring, and a promise the bleeding will slow
Meta has tried to show that it is not simply throwing good money after bad. Earlier this year the company cut about 1,500 people from Reality Labs, a sizable reduction for a single division that signals management’s willingness to trim projects and headcount. Yet executives were explicit with investors that these layoffs do not mean an exit from VR or the metaverse. Instead, they framed the cuts as a way to focus on products that can live inside Meta’s existing apps, a shift that suggests more emphasis on experiences that tie back to Facebook, Instagram and WhatsApp rather than standalone virtual worlds.
Chief executive Mark Zuckerberg has been working to reassure markets that there is an end point to the spending surge. In one investor conversation, the CEO said he expects Reality Labs losses this year to be similar to last year, a blunt admission that 2026 will not bring relief but also a signal that the company believes it is near a peak in cash burn. In a separate discussion, Mark Zuckerberg argued there is an end in sight to the multibillion dollar losses, pointing to layoffs and a push for third party Horizon OS headsets as steps toward a more sustainable model. I read that as a plea for patience rather than a concrete roadmap, because the company is still asking investors to accept another year of roughly $19 billion in losses on faith that the curve will eventually bend.
AI land rush on one side, VR drag on the other
What makes Meta’s VR spending spree more palatable to markets is that the rest of the business is throwing off enormous cash. The company’s ad engine powered a blockbuster quarter, and Meta is now positioning its 2026 capital spending as an AI land rush, with Mark Zuckerberg effectively telling investors that the company must invest aggressively in infrastructure to stay ahead. One recap of the earnings call described how Meta’s AI spending is ramping up, with Jan and other executives emphasizing that the company sees generative models and recommendation systems as the next major growth driver. In that context, Reality Labs looks less like the only big bet and more like one of two massive, parallel platform gambles.
Yet the contrast between the narratives around AI and VR is stark. AI is framed as a near term revenue accelerator that can boost engagement and ad pricing inside Meta’s existing apps, while Reality Labs is still sold as a long term vision that might pay off years down the line. On the same call where Jan and colleagues talked up AI, they also acknowledged that Reality Labs, the division behind Quest headsets and Horizon, remains a drag. I see Meta trying to thread a needle: it wants to keep funding VR at roughly current levels while shifting investor attention to AI, which is easier to justify because it is already embedded in the products that generate most of the company’s revenue.
From VR-first to wearables and mixed reality
Meta is also subtly changing what its metaverse strategy looks like in practice. Instead of doubling down on pure VR, the company is leaning into AI powered wearables such as smart glasses, which can be sold as practical tools rather than portals to a fully virtual world. One earnings recap noted that Meta plans to in favor of these AI powered devices, a shift that aligns with the company’s push to embed assistants and generative features into everyday hardware. That does not mean headsets are going away, but it does suggest that the company sees more near term upside in products that look and feel like regular glasses than in bulky VR rigs.
Inside the company, this evolution is framed as a strategic reshuffling rather than a retreat. One report described how, After losing $20 billion in 2025 alone in the business that Mark Zuckerberg changed the company’s name for, Meta told investors not to expect a quick turnaround, and that the losses came with a strategic reshuffling. That language matters. It signals that Meta is not abandoning the idea of a new computing platform, but is broadening it to include mixed reality, smart glasses and AI assistants that can live on your face or in your pocket, not just in a headset used to hold meetings.
Investors’ patience and the long road ahead
For now, investors seem willing to tolerate the VR losses because Meta’s core business is so profitable and its AI story is compelling. The company’s stock has climbed on strong revenue and guidance, even as Reality Labs remains a heavy weight on the income statement. Yet the messaging from leadership is clear that the pain is not over. One recap of the latest call highlighted that Jan and other executives are effectively telling investors to brace for Reality Labs losses in 2026 that are similar to 2025 levels, and that the division will continue to be a drag even as AI capex climbs. When a management team says, in effect, “Do not expect improvement next year,” it is asking for a rare degree of long term trust.
