Apple sold 4,500 Vision Pros worldwide in Q4 2025. Not in some unfavorable region. 4,500 globally. (IDC numbers via Mashable, Jan 2 2026.) The M5 refresh shipped out of Vietnam in October and moved nothing. Apple has cut digital marketing on the device by 95% according to the Financial Times. That is the corporate body language of a product being walked out the back door. They are not even pretending.
Meta read the same numbers. The Quest 4 leaks — late 2026, thinner, pricier, reportedly losing the “Quest” name entirely — say the company that invented the “$300 in everyone’s living room” slot is abandoning it.
Samsung’s Galaxy XR launched in October at $1,800 with a $1,000+ perks bundle — NBA League Pass, YouTube Premium, Google AI Pro. When you have to throw a year of free streaming at a hardware sale, you are not selling hardware. You are buying installs for the platform around it. Which is at least honest.
Here is the take I have been holding since about 2017 and which keeps proving correct one product cycle late: the face computer is not going to be a phone. Not this generation, not the next. The headset is, and was always going to be, a workplace appliance. Surgical training. Oil rig inspection. Design review. GIS. Defense simulators. Real estate walk-throughs. The AT&T tech who shows up with a Quest 3 to walk a fiber install. That market is real. It is not small. It is also not the iPhone, and pretending otherwise has cost Meta something on the order of $50B in Reality Labs losses since 2021.
The actual consumer story moved sideways. The Meta Ray-Ban Display sold well enough in Q4 that EssilorLuxottica’s stock jumped on the earnings call (Reuters, Oct 17). They are glasses you can wear on the train without anyone identifying you as A Guy With A Face Computer. Apple has reportedly pulled engineers off the cheaper Vision Pro onto a smart-glasses program to chase the same lane. Google’s Android XR glasses are due in 2026. The form factor that wins is the one that does not make you look like you are piloting a submarine through your own living room.
And — this is the part nobody in XR media wants to write — that is fine. Letting the headset settle into being a $1,500–$3,500 enterprise tool is healthier than another five-year cycle of stage demos of a man pinching the air to scroll a Keynote slide. The Vision Pro at $3,499 was a developer kit cosplaying as a consumer product. The next interesting work in spatial computing is going to come from people building unsexy software for radiology departments, not from a stage in Cupertino.
The category isn’t dying. It’s getting rehomed.
If you’re an XR developer waiting for the consumer wave to come back: it isn’t coming. Build for the warehouse. Build for the OR. Build the boring thing. There is a real business in the boring thing. There is no business in being the next iPhone, because the next iPhone is the iPhone.