Why AI glasses sales are exploding (and why VR is collapsing at the same time)

Published on 2026-08-30

Disclosure: Glaiss may earn a commission if you buy through links on this page. It does not change our rating or the price you pay.

In the first quarter of 2026 the world bought 2.25 million display-free AI glasses. A year earlier it had been under a million. That is 167 % growth according to IDC, and Counterpoint, measuring differently, puts it at 210 %.

In the same quarter, VR headsets fell 17 % year on year and 39 % against the previous quarter.

Both things are happening at once, and that simultaneity is the whole story. It is not that a market for "things you wear on your face" is growing. It is that one product is replacing another, and the winner looks nothing like the loser.

First: they are not the same thing, and confusing them explains everything wrong

Most coverage lumps together three products that do not compete with each other:

The 167 % growth belongs to the third row only. Add all three categories into one chart — as many headlines do — and the result means nothing, because you are adding a collapsing category to an exploding one.

The distinction is not academic. It determines what you buy, how much it weighs on your face, and whether you will still be using it in six months.

The numbers, unvarnished

These are the verified figures as of August 2026. There is no public Q2 data from IDC, Counterpoint, Omdia or Canalys, so any headline talking about "all of H1 2026" is estimating.

One detail that gets overlooked: the average selling price is $376. This is not a cheap impulse accessory. It is a deliberate purchase, comparable to a mid-range phone, repeated more than two million times in three months.

The four real reasons

1. The one that does not look like a gadget won

This is the main reason, and it is uncomfortable for the industry because it has nothing to do with technology.

IDC attributes the growth to "fashion and social acceptability". In other words: Ray-Ban Meta sell because they look like Ray-Bans. You do not have to explain to anyone why you are wearing them, you do not have to take them off to talk to someone, and there is no moment where you stop seeing the world.

The Vision Pro did exactly the opposite. Technically it is a feat — the best display ever put in front of a human eye — and commercially it collapsed: from 390,000 units in 2024 to around 45,000 in 2025. That product did not fail on technology. It failed because nobody wants to put a device on their face in front of another person.

The market's conclusion is brutal in its simplicity: the deciding factor was never resolution or latency. It was whether you are embarrassed to walk down the street in them.

2. The assistant is finally useful

In 2023, a voice assistant in glasses was a radio with commands. In 2026, Meta's glasses run Llama 4, the RayNeo iO run Gemini and Rokid lets you choose the model. The practical difference is that you can now look at something and ask what it is, or talk to someone in another language with roughly a second of latency.

That is the function that justifies the spend, and it did not exist when VR headsets were designed. VR was conceived to take you somewhere else; AI glasses were designed to help you where you already are. The second proposition fits what people actually do with their phone.

3. Meta stopped funding the metaverse and started funding this

This is not interpretation, it is in their accounts.

• Reality Labs lost $4.62 billion in the second quarter of 2026, on revenue of just $431 million. The division's accumulated losses are around $88 billion since 2020.
• In April, CFO Susan Li told analysts that VR investment would be reduced significantly and that spending would shift toward wearables.
• Horizon Worlds went into maintenance mode in January 2026, after cuts to the division.
• In 2025, for the first time, glasses revenue ($2.15 bn) overtook Quest ($660 m).

The company that defined the metaverse has reallocated its budget to glasses. When the player holding 69 % of the market moves its money from one category to another, the rest of the industry follows.

Worth understanding: the division is still losing money hand over fist. The glasses sell, but Reality Labs is nowhere near profitable. Meta is buying market share in a category it believes will succeed the phone. That may work out, or it may be the second $88 billion mistake in a row.

4. Chinese brands crushed the price and multiplied the choice

While Meta dominates the display-free segment, the moment a display appears the market is Chinese: Chinese brands control 71 % of the world's AI glasses with a display (Omdia). RayNeo leads AR glasses globally with 41 %, and Viture grew 281 % year on year (Counterpoint).

In China they call it 百镜大战, "the war of a hundred glasses". The effect for a European buyer is that there are models at $299 where two years ago there was one at $1,500. How many of those you can actually buy in Europe, with a warranty that means something, is a separate question.

The counterargument worth taking seriously

The easy thesis is that AI glasses will replace the smartphone. It is worth testing against the people who measure the market.

Francisco Jerónimo of IDC is blunt: "All those ideas that AR and VR would replace smartphones did not happen. They will never happen."

And Ramon Llamas, also of IDC, tempers the growth expectations: "We expect the market to recover somewhat more in 2026, but are we going to see a hockey-stick effect in 2026? I don't think so."

The numbers back them on scale. 13.6 million units forecast for 2026 against more than a billion smartphones a year. This is a market roughly 1 % the size of phones. It is a very successful accessory, not a replacement.

Two more objections belong here:

• The comparison base was tiny. Growing 167 % from under a million units is easier than growing 20 % from a hundred million. The percentage impresses more than the absolute number.
• Nobody knows how many people still use them. What gets measured is what sells, not what is still on someone's face at six months. In a product that lives on battery and habit, that figure is the one that will really decide whether this is a category or a fad — and nobody publishes it.

What this means if you are thinking of buying

Three practical consequences:

Prices will fall, but not this year. IDC forecasts the average price dropping from $376 to $229 by 2030. That is a six-year decline, not a six-month one. Waiting a year saves you little and costs you the use.

The category with the most new models has the worst battery. Display glasses are growing fastest in choice (41.9 % CAGR) and last the least: the RayNeo X3 Pro give an hour and a half. Before novelty pulls you in, read AI glasses with or without a display.

Meta's 69 % share means the ecosystem is where it is. Buying outside Meta is perfectly reasonable — there are better models on weight, on languages or on price — but it means accepting fewer accessories, fewer updates and a smaller community. That is a conscious trade-off, not an oversight.

Frequently asked questions

Why are AI glasses growing while VR collapses, if they are similar products? Because they are not similar. VR covers your eyes and replaces reality; AI glasses look like normal glasses and help you while you live your day. IDC attributes the growth to fashion and social acceptability, not to technology. A product you can wear in front of other people has a far larger market than one you only use alone at home.

How many AI glasses have sold in 2026? 2.25 million display-free units in the first quarter, up 167 % year on year according to IDC. There is no public second-quarter data from the major analyst firms. IDC forecasts 13.6 million units for the full year.

Will AI glasses replace the smartphone? Not in the short term, and probably not in the way it is usually framed. The forecast market for 2026 is 13.6 million units against more than a billion smartphones a year: around 1 %. IDC's Francisco Jerónimo is categorical that the idea of AR and VR replacing the phone will never happen. The reasonable view is a phone accessory, not a successor.

Was the Apple Vision Pro a failure? In units, yes: from around 390,000 in 2024 to about 45,000 in 2025. Technically it is an extraordinary product, but it proved the category's problem was not display quality — it was putting a visible device on your face. That lesson is precisely what explains the success of glasses that look like glasses.

Is Meta making money on glasses? No. Reality Labs, the division that builds them, lost $4.62 billion in the second quarter of 2026 alone on $431 million of revenue, and has accumulated roughly $88 billion in losses since 2020. What did happen in 2025 is that glasses revenue overtook Quest for the first time, which is why Meta has shifted its investment toward them.

Is it worth waiting for prices to drop? Not much. IDC forecasts the average price falling from $376 to $229, but by 2030. Over a one-year horizon the expected drop is small, whereas new models certainly will arrive: if anything justifies waiting a few months it is a specific launch, not the price.

Sources: IDC (Q1 2026 shipments, 2026–2030 forecasts, average price), Counterpoint Research (Q1 2026, AR share), Omdia (Chinese share of display glasses), Meta Q2 2026 financial results, EssilorLuxottica (2025 units).