Are AI glasses the replacement for the smartphone?
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.
This is analysis, not a review. We are arguing a case with data here. If you want to know what to buy today, start with the three types of smart glasses.
The question has been going around for two years, and it is almost always asked badly.
"Replace the smartphone" sounds like one device disappearing and another taking its place, the way DVD took over from VHS. Framed that way, the answer is easy and boring: no. Not one of the thirteen models in our catalogue works without a paired phone. Not one. IDC says it plainly in its 2026 analysis: Meta's glasses depend on a phone for full functionality, and Google's and Samsung's will be born deeply integrated into Android XR. Technically speaking, they are a phone accessory.
And the volume comparison is almost comic. In 2026 the world will buy 13.6 million display-less AI glasses. And just over a billion smartphones. Glasses are 1.3 % of the market they are supposedly replacing.
The article could end there. But that would be stopping at the surface, because that is not the interesting question.
The right question
The smartphone does not sell a device. It sells the moment you take it out of your pocket: check the time, answer a message, find the way, take a photo, ask something. Fifty, a hundred, a hundred and fifty times a day.
That is the product. The six-inch pane of glass is just the packaging.
So the right question is: how many of those moments can be settled without taking it out? Because every one solved by voice, or by a line of text on a lens, is a moment the phone loses for good — even though the phone stays in the pocket and charges every night.
That is our thesis: AI glasses are not going to replace the phone. They are going to retire it as the primary screen. It is a substitution of attention, not of hardware, and it is exactly what happened once before.
The precedent nobody remembers correctly
The phone did not kill the PC. Computers still sell in 2026, and anyone who writes code, edits video or keeps accounts still sits down at one.
What the phone killed was the reason to sit down. Checking email, looking at the bank, buying something, arguing with strangers — all of it left the desk and never came back. The PC did not disappear; it specialised. And today nobody would call the PC the dominant platform of consumer computing, even though it still ships by the tens of millions.
That is the pattern. One format does not replace another: it takes the frequent uses and leaves the deep ones. The phone will keep the deep ones — writing at length, editing, watching a film, anything needing two hands and full attention. Glasses will take the frequent ones.
The numbers behind the thesis
This is where 2026 gets interesting, because both curves are moving at once and in opposite directions.
AI glasses, in the best year they have ever had:
- +167 % year on year in Q1, with 2.25 million units in three months (IDC).
- 13.6 million units and $5.1 billion forecast for the full year.
- More than 7 million Ray-Ban and Oakley Meta sold in 2025 alone, more than triple all of 2023 and 2024 combined, per EssilorLuxottica.
- Meta holds 69.2 % of the market. This is not a lab experiment; it is a consumer product with a clear leader.
- Meta and EssilorLuxottica are in talks to scale production to 20 million units a year, nearly triple 2025.
Smartphones, in the worst year of theirs:
- −16.7 % in units in 2026. IDC calls it the steepest annual contraction the industry has ever recorded.
- Average selling price up 27.6 % to $581, because NAND and DRAM costs have risen over 300 % year on year.
- The shortage is expected to last until 2028 at the earliest, and it is erasing the budget tier of the market.
Be careful about what this means and what it doesn't. Glasses are not causing the phone decline. A memory crisis is, and it has nothing to do with them. But the effect matters: phones are getting more expensive and less accessible exactly as a category appears whose units are getting cheaper — IDC expects the average price of AI glasses to fall from $376 today to $229 by 2030, a 40 % drop.
One market cools and gets pricier while the other heats up and gets cheaper. That does not prove substitution, but it is precisely the terrain on which substitutions happen.
And the argument that isn't a number
Mark Zuckerberg said it on Meta's earnings call of 3 August 2025, answering this exact question: anyone without AI glasses will be at "a pretty significant cognitive disadvantage". He compared it to needing glasses and not wearing them.
He is an interested party, obviously: he sells glasses and holds 69 % of the market. But the argument is not stupid. Taking the phone out has a cost. It costs you the interruption, looking down, unlocking, finding the app and coming back. That cost is high enough that most people simply don't look up things they would benefit from looking up.
Driving that cost to zero is a change of category, not an improvement. It is the difference between having access to something and actually using it.
The five obstacles, unvarnished
None of the above matters if these aren't solved. And today they aren't.
1. Total dependence on the phone. No pair in the catalogue works alone. They carry no SIM, and neither the connectivity nor the battery would support one. While a phone is mandatory, glasses are a peripheral, and a peripheral replaces nothing. It is obstacle number one and no brand has announced it will break it.
2. Battery. The hardest number in the catalogue: models with a camera and a colour display last between ninety minutes and six hours. The RayNeo X3 Pro, the most technically advanced of the lot, manages ninety minutes. Only the ones that drop the camera and use a monochrome display reach 48 hours, like the Even Realities G2 or the RayNeo iO. A device meant to be on all day cannot last an hour and a half.
3. How you type. The phone won because someone put a keyboard on the screen. On glasses you write with your voice, and voice does not work in an office, on a train, or when what you are writing is private. Neural rings and wristbands are on the way, but today text input is an unsolved problem, and without it there is a long list of things only the phone can do.
4. Manufacturing waveguides. The display in AI glasses is made with optical waveguides, and producing them at volume is the industry's real bottleneck. That is not opinion: on 16 June 2026 Meta and EssilorLuxottica signed a deal with Applied Materials specifically to solve it. Until it is solved, 20 million units a year is an intention, not a capacity.
5. People actually wanting to wear them. This is the soft obstacle, and the most underestimated. Google Glass did not fail on technology; it failed because people who wore them came across badly. The reason Ray-Ban Meta works is precisely that it does not look like a device — IDC credits its growth to fashion and social acceptability, not to specifications. The moment a display forces them to look like technology, that advantage can evaporate.
What would have to happen for us to be right
Rather than hide behind an unfalsifiable thesis, here is what we will be watching. If it doesn't happen, we were wrong.
- Glasses with their own connectivity, no phone required, even with reduced functions.
- A full day of battery with a camera and a colour display. Today the catalogue maximum for that combination is six hours.
- A way to write that isn't voice, and that people actually use.
- Average price below the $229 IDC forecasts for 2030. At $376 they are a second device; under $200 they start being the first.
- Annual sales moving from tens of millions to hundreds. IDC's 2030 forecasts — 27.3 million display-less, 12.2 million optical see-through, 10.4 million mixed reality — add up to around 50 million. That is still 4 % of the phone market. IDC's own optimistic scenario does not describe a substitution.
That last point deserves reading twice, because it is the one that stings. Our thesis is that glasses take the attention, not the units — but if they are still 4 % of the installed base in 2030, the attention won't have moved much either.
So what do you do today?
If you came here wondering whether to buy a pair so you can stop using your phone: you are not going to stop. You will take it out less, which is a different thing and already quite a lot.
With Ray-Ban Meta or Oakley Meta HSTN you stop reaching for the phone to take a photo, play music, take a call or ask something. With a display model, also to read a message or follow a route. Everything else still happens on the screen in your pocket.
Whether that is worth it depends on how many times a day you pull out your phone for something trivial. If the answer is "a lot", the purchase makes sense today. If it's "rarely", wait: in two years they will be better and cheaper, and none of the five obstacles above gets solved by buying now.
Frequently asked questions
Can AI glasses work without a phone? Not today. None of the thirteen models we track works without a paired phone: they carry no SIM, and neither connectivity nor battery would support it. IDC points this out explicitly in its 2026 analysis, and it is the main technical argument against the idea that they replace the smartphone.
How many AI glasses sell compared to phones? In 2026 the forecast is 13.6 million display-less AI glasses against just over a billion smartphones — roughly 1.3 %. Even in IDC's 2030 forecast, adding all three families, around 50 million units is still about 4 % of the phone market.
Is the smartphone going to disappear? No data supports that, and we are not arguing it. What is happening is that glasses are taking the fast, frequent uses — a question, a photo, a notification, a route — and leaving the phone the ones that need two hands and full attention. It is the same pattern by which the computer stopped being the primary screen without ceasing to sell.
Why are phone sales falling in 2026? Because of a memory shortage, not because of glasses. NAND and DRAM costs are up more than 300 % year on year, the average phone price is up 27.6 % to $581 and units are down 16.7 %, the largest annual contraction on record. IDC expects the shortage to last until 2028 at the earliest.
What is missing before glasses could replace a phone? Five things, none of them solved: their own connectivity without a phone, a full day of battery with camera and colour display, a way to write that isn't voice, industrial capacity to make waveguides at volume, and people wanting to wear them once they stop looking like ordinary glasses.
Is it worth buying a pair now if that's what I'm after? Only if your goal is taking the phone out less, not putting it away. For hands-free photos, music, calls and questions, the display-less category already does the job well at €419–549. For anything that needs typing, long reading or real work, the phone remains irreplaceable.
What exactly did Zuckerberg say? On Meta's earnings call of 3 August 2025 he said anyone without AI glasses would be at "a pretty significant cognitive disadvantage", comparing it to needing glasses and not wearing them. He gave no date for smartphone replacement, and it is worth remembering Meta controls 69 % of this market.
Sources: IDC — Smart Glasses Market 2026: XR Is Rewriting the Rules (shipments, share and forecasts to 2030) and Smartphone Shipments Set for Record 16.7% Drop in 2026 (units, average price and the memory crisis). EssilorLuxottica full-year 2025 and Q2 2026 results. Meta Q2 2025 earnings call. Meta–EssilorLuxottica–Applied Materials agreement, 16 June 2026. Weight, battery and field-of-view figures from our own verified model pages.