For a decade, "smart glasses" was a punchline. Google Glass arrived in 2013, hit a wall of privacy backlash — wearers were nicknamed "Glassholes" — and retreated. The heavy virtual-reality headsets that followed sold to gamers and developers but never became the everyday device their makers promised. So the past year is worth noticing. In September 2025, Meta put a small color display and a muscle-reading wristband on a pair of Ray-Ban glasses and began selling them in stores [source: Meta, 2025]. By the second half of 2025, one market tracker estimated that global smart glasses shipments had grown 139% year over year [source: Counterpoint Research, 2026]. The category everyone had written off was suddenly the fastest-moving corner of consumer hardware.
That momentum arrived wrapped in a familiar marketing fog, and an honest picture is more useful than either the hype or the old cynicism. Some of these devices are real products you can buy today; others are polished prototypes or dates on a roadmap. Some are lightweight glasses; others are headsets you strap to your face. And the numbers describing the "market" are estimates that disagree with one another. This article sorts what shipped from what was merely announced, separates three very different products that share the name "smart glasses," and weighs the genuine promise against the unresolved problems of privacy, battery, and the still-missing killer app. It is not investment advice or an endorsement of any company.
Table of Contents
- Why smart glasses became a 2026 story
- Three different things called "smart glasses"
- What actually shipped versus what was only announced
- The numbers, and why to read them as estimates
- The platform war: Meta, Android XR, and Apple
- The always-on camera problem
- The honest limits: killer app, battery, and price
- The bottom line: what to watch
Why smart glasses became a 2026 story
The turning point was not a single breakthrough but a shift in strategy. Instead of strapping a computer to your face, Meta and its eyewear partner EssilorLuxottica put modest technology — a camera, speakers, microphones, and a voice assistant — into a normal-looking pair of Ray-Ban frames. That first Ray-Ban Meta generation had no display at all, and that was the point: it looked like eyewear, not gear. It sold well enough that EssilorLuxottica moved to ramp annual production toward 10 million units by the end of 2026, and in January 2026 Bloomberg reported the two companies were discussing doubling that to 20 million or more [source: Bloomberg, 2026]. By the company's own telling, smart glasses now account for more than a third of EssilorLuxottica's revenue growth [source: Bloomberg, 2026].
At Meta Connect in September 2025, the company pushed further with three products at once: a second-generation Ray-Ban Meta starting around $379, a sportier Oakley Meta Vanguard at $499, and — the headline — the Meta Ray-Ban Display at $799 [source: Engadget, 2025]. The Display is the pivotal one, because it adds what the earlier glasses deliberately left out: a small, full-color screen inside the right lens, paired with a wristband that reads the electrical signals in your muscles so you can control it with tiny finger movements [source: Meta, 2025]. That combination — an ordinary-looking frame that can now show you something — turned a niche accessory into a story about the next computing platform.
Three different things called "smart glasses"
The single most useful distinction here is that "smart glasses" now covers three product categories that differ enormously in what they do, what they cost, and how far along they are. Blur them together and every claim gets confusing.
Audio-and-camera glasses
The best-selling category has no display. Ray-Ban Meta and Oakley Meta Vanguard let you take photos and video, listen to audio, and talk to an AI assistant — but they show you nothing. Everything happens through sound and capture. These drive the shipment numbers, because they are the lightest, cheapest, and most socially normal to wear.
Heads-up display glasses
The second category adds a screen. The Meta Ray-Ban Display puts a small color heads-up display (HUD) in one lens, off to the side so it does not block your view, showing texts, navigation, or an AI reply [source: Meta, 2025]. A separate cluster of products from companies such as XREAL, Viture, and RayNeo uses optical see-through displays, often tethered to a phone or console for watching video. These are true display glasses, but the display is small and monocular, not an immersive field of holograms.
Full mixed-reality headsets
The third category is not glasses at all. Apple Vision Pro and Samsung Galaxy XR are headsets worn over the eyes, with high-resolution internal screens that reproduce the world through cameras. Apple markets its device as "spatial computing," and it is powerful — the 2025 refresh added Apple's M5 chip and a more comfortable band, still priced at $3,499 [source: Apple, 2025]. But it weighs on your face, costs as much as a laptop, and is a fundamentally different thing from a pair of Ray-Bans. Meta's own true-AR prototype, Orion, sits in a fourth bucket entirely: a device Meta explicitly says it will not sell.
What actually shipped versus what was only announced
The discipline that matters most here is the line between a product on a shelf and a promise in a keynote. Several genuinely shipped in 2025. The Meta Ray-Ban Display went on sale September 30, 2025, at a limited set of US retailers, with the UK, France, Italy, and Canada slated for early 2026 [source: Meta, 2025]. Samsung's Galaxy XR headset went on sale October 21, 2025, at $1,799.99 [source: Samsung, 2025]. Apple's M5 Vision Pro reached general availability on October 22, 2025 [source: Apple, 2025]. These you can buy.
Several equally famous devices are not for sale. Meta's Orion, unveiled in September 2024, is the clearest example: it displays life-size holograms across a wide field of view, but Meta states plainly that "Orion won't make its way into the hands of consumers," describing it as a prototype for internal and select external testing rather than something you can purchase [source: Meta, 2024]. Reporting has put its build cost on the order of $10,000 per unit, which explains why. Snap's consumer AR glasses, branded "Specs," are another: Snap said in June 2025 they would arrive in 2026 and has since shown them publicly, but as of mid-2026 they are a planned launch, not a shipping product [source: Bloomberg, 2025]. And while Google's Android XR platform is real and runs on the Galaxy XR headset, the lightweight Android XR glasses Google has demonstrated remain a developer-stage promise, not a device on sale.
Why insist on this line? Because the most exciting demonstrations — wide-field holographic AR, all-day glasses that replace your phone — tend to belong to devices that have not shipped, while the products you can actually buy do less. Judging the category by its prototypes overstates where it is; judging it only by today's shelves understates where it is heading. Both readings need the label attached.
The numbers, and why to read them as estimates
The growth figures are striking, and they are estimates from private research firms that measure the market differently. Counterpoint Research estimated that global smart glasses shipments grew 139% year over year in the second half of 2025, and put Meta's share at about 82% for that period [source: Counterpoint Research, 2026]. IDC, using its own methodology, forecast full-year 2026 shipments of roughly 13.6 million units, rising to about 27.3 million by 2030, and pegged Meta's share at 69.2% in the first quarter of 2026 [source: IDC, 2026]. The two firms disagree on Meta's exact share — 82% versus 69.2% — not because one is wrong, but because they cover different periods and define the category differently. Treat any single market-share or shipment number as one estimate among several.
Two patterns survive the disagreement. First, the market is small but growing fast: IDC's 13.6 million units for 2026 is a fraction of the roughly 1.2 billion smartphones sold each year, so "fastest-growing" and "still niche" are both true at once. Second, the display-glasses slice is the fast riser within the fast riser: IDC expects optical see-through display glasses to grow from about 3 million units in 2026 to 12.2 million by 2030 [source: IDC, 2026], and Counterpoint separately estimated that AR (waveguide-based) smart glasses shipments jumped 148% year over year in the second half of 2025 [source: Counterpoint Research, 2026]. IDC also expects the average selling price to drift down from around $376 today toward roughly $229 by 2030 [source: IDC, 2026] — a sign of mainstreaming, and of thinner margins that could test how durable the boom is.
The platform war: Meta, Android XR, and Apple
Underneath the hardware, three companies are trying to own the software layer, and their bets differ. Meta's is the volume play: cheap-to-mainstream glasses sold through EssilorLuxottica's retail reach, with the Ray-Ban Display as the bridge toward the full-AR future its Orion prototype hints at. Meta frames these as "AI glasses" that could eventually rival the smartphone — a company claim no independent evidence yet supports, better read as ambition than fact.
Google's bet is the platform it knows best. In December 2024, Google announced Android XR, an operating system built with Samsung and Qualcomm for both headsets and glasses, with its Gemini AI assistant woven in [source: Google, 2024]. The first device to run it, Samsung's Galaxy XR, is a headset, but Google has been explicit that the same platform is meant to power lightweight glasses too — an attempt to repeat the Android strategy of supplying the software while partners build the hardware.
Apple's bet is the premium, tightly integrated headset. The M5 Vision Pro is the most powerful device of the three, and Apple calls the experience "spatial computing" rather than AR or VR [source: Apple, 2025]. But at $3,499 and worn over the eyes, it targets a different user than a $379 pair of Ray-Bans. Widely reported to be working on lighter glasses of its own, Apple has shipped only the headset so far. The contest is not one race but several overlapping ones, and it is far from settled which form factor, if any, becomes the one people wear all day.
The always-on camera problem
The feature that makes these glasses useful — a camera and microphone always on your face — is also their central controversy, and it deserves both sides. The concern is real: a camera worn at eye level, indistinguishable from ordinary glasses, means people around you cannot easily tell when they are being recorded. Critics argue bystanders have little control over being filmed or over how the footage is used, and two US lawsuits have alleged Meta misled consumers about the glasses' privacy implications [source: Fortune, 2026]. European regulators, including data-protection authorities in Italy and Ireland, have questioned since 2021 whether a small indicator light is enough to alert people that a camera is running [source: Fortune, 2026].
The indicator light is where the debate sharpens. Meta's glasses include an LED that lights up when capturing, its answer to the bystander problem. But critics note the light is easy to miss in daylight, that some users have found ways to defeat it, and — most pointedly — that in mid-2026 Meta was reported to be testing an always-on "super-sensing" AI mode in which the LED would not activate, making covert recording harder to detect [source: Fortune, 2026]. Meta, for its part, has also tightened some privacy safeguards and argues the glasses are designed with capture indicators and user controls.
There is a genuine trade-off underneath, not just a villain. The same hands-free capture that worries a bystander can help a low-vision user navigate, let a worker document a repair, or translate a foreign menu in real time. The Google Glass backlash of 2013 showed that social acceptability, not just engineering, decides whether a face-worn camera survives contact with the public. Whether this generation clears that bar — through clearer signals, stronger norms, or regulation — is one of the open questions the technology has not answered.
The honest limits: killer app, battery, and price
Strip away the platform ambitions and a plain question remains: what are these glasses for? Today the answer is a set of useful-but-modest jobs — capturing a moment, taking a call, hearing directions, asking an AI a question, showing a notification. Those are real conveniences. But none is yet the killer app that makes glasses a device you cannot live without, the way messaging and the camera made the smartphone indispensable. The industry's boldest claim — that AI glasses will replace the phone — is a marketing thesis, not a demonstrated fact, and should be labeled that way.
The hardware imposes its own ceilings. A display bright enough to read outdoors, a camera, and a wireless radio all draw power from a battery that has to fit inside a slim temple arm, so runtime is measured in hours, not days. The heads-up display in devices like the Ray-Ban Display is small and monocular — a window off to the side, not an immersive overlay — a deliberate engineering compromise, but far from the holographic vision the prototypes sell. And price still sorts the market: the display glasses carrying the future-defining features cost $799 and up, while the volume is driven by cheaper, display-free models. Even IDC's forecast of falling prices is a multi-year story, not a today one. None of this means the category is failing; it means the useful version and the visionary version are not yet the same product, and the distance between them is measured in years of battery chemistry, optics, and social negotiation.
The bottom line: what to watch
Smart glasses in 2026 are a real, fast-growing category where demonstrations run well ahead of what you can buy. The genuine shift is that lightweight glasses, not bulky headsets, have become the mainstream form factor. Yet the flashiest capabilities belong to prototypes like Orion that are not for sale, the market numbers are estimates that disagree, the always-on camera has not resolved its privacy problem, and no killer app has made glasses indispensable.
So watch the milestones that move the field, not the ones that make headlines. Does a display-glasses product find a job people do every day, not just a demo? Do IDC's and Counterpoint's shipment estimates keep climbing once the novelty fades, and does the average price really fall? Does Android XR reach lightweight glasses, and does Apple ship anything lighter than a headset? And can industry, regulators, and the public settle the terms under which a face-worn camera is socially acceptable? For now the honest promise is smaller than the pitch and still meaningful: the glasses on the shelf can take a photo, read you a message, and answer a question — and for the first time in a decade, enough people want that to make the rest worth watching.