Right now, today, you can reportedly walk into an Apple Store and hand over $3,699 for a Vision Pro with an M5 chip. Apple will sell you one. What it won't tell you is that behind that device sits a product roadmap that now contains, by the company's own reported reckoning, exactly zero successors.
That's the trap. The M5 Vision Pro is not a discontinued clearance item — it's on active sale at flagship money. But the platform underneath it has been quietly hollowed out, and the people who bought in are the ones left holding the bag.
Last issue, in the Ternus Reset item, we flagged the Vision Pro cancellation as a rumor — a strategy pivot that hadn't yet been confirmed by anything you could put on record. This week it's reportedly confirmed. AppleInsider reports Apple has laid off at least 60 employees tied to its VR team, the Apple Vision Group, and similar roles. 9to5Mac also reported the number, citing AppleInsider. The category, per that reporting, has reportedly been put "on ice" by incoming CEO John Ternus.
So the rumor is now a headcount. And a headcount is the clearest signal a company ever sends.
So the rumor is now a headcount. And a headcount is the clearest signal a company ever sends.
Here's the scale of what was actually built. Apple sold roughly 600,000 Vision Pros in total across the device's entire life. For a company that moves that many iPhones before lunch, that's a rounding error — and MacRumors reports Apple also saw an unusually high return rate, far exceeding any other modern Apple product.
Think about that combination. A tiny installed base, and an unusually large chunk of even that tiny base sent the thing back. This was never a mass platform. It was an expensive experiment, and the experiment is over.
The most damning confirmation didn't come from a leaker. It came from Ming-Chi Kuo. A year ago, Kuo tracked seven head-mounted wearables in various stages of development inside Apple. As of his June update, only two survive — and neither is a headset.
What's left: display-less AI glasses shipping in 2027, and waveguide-powered AR glasses that have now slipped to 2029. The M5 Vision Pro already shipped. Every other headset — including the lighter, cheaper "Vision Air" that was supposed to fix the price and weight problems — is reportedly gone from the roadmap.
Kuo was explicit about who signed off: "The major overhaul was signed off by Apple's next CEO, John Ternus."
That's the whole story in one sentence. The person about to run Apple looked at spatial computing and chose to keep two products, both glasses, one of which won't even have a display, the other of which is three years out.
Now hold that next to what Apple's marketing department is still saying. Two months ago, Steve Sinclair — who leads product marketing for visionOS — was asked directly about reports that Vision Pro had been put on ice.
His answer: "It's still early innings for spatial computing... I would hope that updates like visionOS 27 and all the things that we're adding... serves as evidence that we are invested in this platform."
Read that again with the layoffs in mind. The "team" being cited as evidence of investment is, in part, the same team that just got let go. "Early innings" is not a roadmap. It's a talking point.
And it's a talking point with a tell. The former Vision Pro chief, Mike Rockwell, has been running the Siri revamp since March 2025. MacRumors reports that some ex-Vision Pro staff have been redistributed to Siri. So the shrinking group still nominally tending visionOS is, functionally, moonlighting on Apple's actual priority — AI.
That's what "invested in this platform" means in practice: fewer people, splitting their attention, keeping the lights on.
None of this makes the hardware bad. The M5 refresh was reportedly real — 10 percent more rendered pixels, a 120Hz refresh rate, about 30 extra minutes of battery. The problem was never that the Vision Pro couldn't render. It's over 1.3 pounds, and even the nicer Dual Knit Band couldn't make it comfortable enough to wear for hours.
Apple knew the fixes. The Vision Air was supposed to deliver them. Apple reportedly killed it.
So let's talk about who actually gets hurt here, because it isn't the tech press.
It's the enterprise buyer. Apple's own executives — Ternus included — pointed to enterprise and medicine as the growing use cases, the serious, sticky, longevity-justifies-the-price customers. Those buyers were sold durability. They deployed $3,699 headsets into training programs and clinical workflows on the premise that this was a platform Apple would build on for years.
Now those same buyers own a $3,699 appliance with a support clock ticking and no next model to migrate to. There's no upgrade path. There's a dead end with a warranty.
And it's the developer. If you built for visionOS, understand what you're building for now: a maintenance-mode operating system with a shrinking team, a ~600,000-unit ceiling, and no hardware successor to expand that ceiling. visionOS 27 will ship with its Siri and Visual Intelligence upgrades, and it will look like momentum. It isn't. It's the platform being kept warm while the people who ran it move to glasses.
The cruel detail for developers: Apple has reportedly not even been able to reuse Vision Pro's core technology in the glasses, because that tech draws too much power for a smaller device. So the spatial-computing knowledge you invested in doesn't cleanly transfer to the thing Apple is actually building next. You're not early on a curve. You're on an island.
Here's the timeline that should reframe every purchase decision. The next spatial bet from Apple isn't a headset. The 2027 glasses reportedly won't have a display at all — they're a Ray-Ban Meta-style AI wearable with cameras and Siri. The first Apple device that actually puts AR imagery in front of your eyes again is reportedly the waveguide model, and that's now 2029.
So if you want an Apple face-computer with a real display, and you don't want an orphaned one, your options are: buy the dead-end Vision Pro today, or wait roughly three years for a product category Apple hasn't shown anyone yet.
That's the verdict. Apple didn't so much kill the Vision Pro as strand the people who bought one — 600,000 early adopters, some enterprise fleets, a developer community — on a platform the company itself has stopped building for while still charging flagship prices to add more people to it.
Should you buy the M5 Vision Pro at $3,699? Only if you want exactly what it does today and never expect it to become more, because it won't. Should you build new for visionOS? Treat it as maintenance, not a bet.
And if a marketer tells you it's still early innings, ask them who's left on the team.