The obituaries write themselves. Meta spent more than 80 billion dollars chasing a virtual world, and this year it is walking Horizon Worlds away from the very headsets it built that world for.
The short version: no, the metaverse is not dead, but Meta’s headset-first version of it is being retired. New Horizon Worlds development is moving to mobile and AI, after Reality Labs losses topped 80 billion dollars since 2020 and the VR app reached only about 200,000 active users. The broad idea of persistent shared digital spaces is alive; it is simply moving to the device three billion people already carry.
What is actually happening to Horizon Worlds?
Separate the news from the noise. Meta is shifting Horizon Worlds development decisively toward mobile, and it pulled back from its harshest plan: after first signaling a VR shutdown, the company decided not to kill VR support outright, keeping existing VR-compatible worlds accessible while placing VR in maintenance mode. The worlds you already have keep working. The company’s money, attention, and new features are moving elsewhere.
Why is Meta pulling back from VR?
The math explains it. Reality Labs has lost more than 80 billion dollars since 2020, and Horizon Worlds reached only around 200,000 active users, modest against an investment that size. Meta has redirected spending toward AI, smart glasses, and wearables, the categories where it is actually seeing people show up. When a bet that large is not paying off, discipline is not surrender. It is what keeps the next bet alive.
A metaverse you reach through the phone in your pocket is not a smaller dream. It is a reachable one.
Is the metaverse dead, then?
No. The vision that struggled was a narrow one: that millions of people would strap on a 500 dollar headset to socialize in a digital room. That specific idea hit a wall. The broader one, persistent shared digital spaces tied to real identity and commerce, is in better shape than the headlines suggest. Analysts still project metaverse-adjacent spending in the trillions by the end of the decade, and roughly a third of organizations expect metaverse-ready products or services on their roadmaps. What is really happening is a correction toward accessibility, and that is where AI enters the story. Pair the reach of a phone with AI that can generate environments, translate conversations, and personalize a space in real time, and the friction that kept ordinary users out begins to fall away.
What does it mean for creators and small businesses?
If you build for immersive platforms, or were thinking about it, the lesson is freeing: design for the device people already own, not the one you wish they owned. Mobile-first spatial experiences, lightweight AR through a phone camera, and AI-driven personalization are where the next wave of opportunity sits. The expensive, headset-only build is now the riskier wager. It also reinforces a trend we have followed closely, that open, cross-device approaches gain ground as walled VR gardens stumble. See our coverage of an open metaverse browser built for small businesses, a mixed-reality browser engine that cuts AR costs for small teams, and the push for open standards from the Open Metaverse Academic Alliance.
So no, this is not a funeral. It is Meta conceding that the road to a shared digital world runs through phones and AI, not through a headset most people never bought. For anyone building rather than mourning, that is good news, because the platform of the future just got dramatically cheaper to build for. Do you think it belongs on the phone, or did the headset crowd have it right all along? I would like to hear the case either way.
Frequently asked questions
Is Meta shutting down Horizon Worlds?
Not entirely. Meta is moving new development to mobile and placing VR in maintenance mode. After initially signaling a VR shutdown, it decided to keep existing VR-compatible worlds accessible.
Is the metaverse dead in 2026?
The headset-first version struggled, but the broader concept of persistent shared digital spaces is not dead. It is shifting toward phones and AI, where reach and low friction make adoption far more realistic.
Why did Meta’s metaverse lose so much money?
Reality Labs has lost more than 80 billion dollars since 2020 while Horizon Worlds reached only about 200,000 active users, a mismatch that pushed Meta to redirect investment toward AI and wearables.
What should small businesses build for now?
Focus on mobile-first and phone-based AR experiences enhanced with AI, rather than headset-only projects. The audience, and the lower cost, are on the phone.
{“@context”:”https://schema.org”,”@type”:”FAQPage”,”mainEntity”:[{“@type”:”Question”,”name”:”Is Meta shutting down Horizon Worlds?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”Not entirely. Meta is moving new development to mobile and placing VR in maintenance mode. After initially signaling a VR shutdown, it decided to keep existing VR-compatible worlds accessible.”}},{“@type”:”Question”,”name”:”Is the metaverse dead in 2026?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”The headset-first version struggled, but the broader concept of persistent shared digital spaces is not dead. It is shifting toward phones and AI, where reach and low friction make adoption far more realistic.”}},{“@type”:”Question”,”name”:”Why did Meta’s metaverse lose so much money?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”Reality Labs has lost more than 80 billion dollars since 2020 while Horizon Worlds reached only about 200,000 active users, a mismatch that pushed Meta to redirect investment toward AI and wearables.”}},{“@type”:”Question”,”name”:”What should small businesses build for now?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”Focus on mobile-first and phone-based AR experiences enhanced with AI, rather than headset-only projects. The audience, and the lower cost, are on the phone.”}}]}
Holoearth Metaverse Shutdown June 2026: What It Means
June 17, 2026 @ 8:08 am
[…] is not a problem unique to Hololive. Meta spent over $80 billion on Horizon Worlds and pivoted away from VR-first to mobile because that is where users actually were. The pattern […]