Meta Sold 7 Million Smart Glasses Last Year. It Also Lost $19 Billion.

Eighty-three point five billion dollars.

That is the cumulative operating loss Meta's Reality Labs division has posted since 2020. In 2025 alone, Reality Labs lost $19.19 billion on $2.21 billion in revenue. For every dollar Reality Labs earned, it spent roughly $9.68.

But something happened in 2025 that has not happened before in six years of billion-dollar bonfires: people actually bought the product. A lot of people.

7 Million Pairs

EssilorLuxottica, the Italian-French eyewear conglomerate that manufactures Ray-Ban Meta glasses, reported selling over 7 million smart glasses in 2025 during its Q4 earnings call in February 2026. Combined with roughly 2 million units sold from the October 2023 launch through early 2025, the installed base now sits at approximately 9 million pairs. North American wholesale growth was described as "exponential."

Seven million units in a single year is a serious number. It puts Meta's smart glasses ahead of several iconic product launches at the same stage. Here is the comparison:

Device	Maker	Year 1 Sales	Year 2 Sales	YoY Growth
iPhone	Apple	6.1M (FY2008)	11.3M (FY2009)	1.85x
iPad	Apple	~14.8M (FY2010)	~28.7M (FY2011)	1.94x
Apple Watch	Apple	~11.6M (2015)	~12.5M (2016)	1.08x
Ray-Ban Meta	Meta/EL	~2M (Oct 2023-early 2025)	7M (2025)	3.5x
Google Glass	Google	~10K (2013)	Canceled	N/A

Year-over-year growth of 3.5x is faster than iPhone (1.85x) and iPad (1.94x) managed in their equivalent periods. Apple Watch barely grew at all in Year 2 (1.08x). Google Glass never made it past an estimated 10,000 Explorer Edition units before the program was shelved.

A caveat: the "Year 1" comparison is imprecise. Meta's glasses launched mid-cycle (October 2023), so the first 2 million were spread over roughly 16 months, not 12. iPhone's Year 1 was a full fiscal year starting from its June 2007 debut. Apple Watch year-1 sales are analyst estimates, since Apple never disclosed unit volumes. The directional signal is still significant: a face computer is selling at a pace that matches the early trajectories of the most successful consumer electronics launches in history.

$1,698 Per Pair

Here is the math Meta probably does not want you to run.

Reality Labs operating loss in 2025: $19.19 billion. Revenue: $2.21 billion. Net burn: $16.98 billion. CFO Susan Li stated on the Q4 2025 follow-up call that Reality Labs would allocate approximately 70% of its 2026 operating expenses to wearables and 30% to VR and Horizon. Using that 70/30 split as a proxy for 2025, approximately $11.89 billion of the net burn went to the wearables program.

Divide $11.89 billion by 7 million glasses sold: $1,698 per pair in R&D subsidy. Even at a more conservative 50/50 split (if wearables received less focus in 2025 than the 2026 guidance implies), the number is $1,213 per pair. The floor is still four times the retail price.

Ray-Ban Meta glasses retail for $299 to $799, depending on lens configuration. Meta is spending between two and six times the retail price of each unit on the R&D required to build and improve them. Put differently: if Meta handed every buyer a stack of $100 bills instead of a pair of glasses, it would cost the company less money.

This is not a gotcha. Platform bets always lose money early. Amazon's Kindle, Sony's PlayStation 3, and Microsoft's Xbox all shipped hardware at a loss to build an ecosystem. What makes the Meta number unusual is the scale of the subsidy relative to the product's price point. Sony lost an estimated $240 per PS3 on a $499 device (a 48% subsidy ratio). Meta's implicit subsidy ratio is 567%.

Where the $83.5 Billion Went

Reality Labs has lost money in progressively larger amounts every year since Meta started reporting it as a separate segment:

Year	RL Revenue	RL Operating Loss	Cumulative Loss
2020	$1.14B	$6.62B	$6.62B
2021	$2.27B	$10.19B	$16.81B
2022	$2.16B	$13.71B	$30.52B
2023	$1.90B	$16.12B	$46.64B
2024	$2.15B	$17.73B	$64.37B
2025	$2.21B	$19.19B	$83.55B

Revenue has been essentially flat at $2 billion per year since 2021, while losses have nearly tripled. Most of that revenue comes from Quest headset hardware and software sales. Meta does not break out glasses-specific revenue, so the actual financial contribution of 7 million pairs of smart glasses to that $2.21 billion is unknown.

The 70/30 pivot announced in January 2026 is the first public admission that the metaverse VR bet, as originally pitched during the October 2021 rebrand, is being downsized. Li described VR consumer adoption as being "on a slower growth path than wearables." Multiple VR game studios, including Sanzaru Games, Twisted Pixel, and Armature Studio, were laid off. Zuckerberg's 2021 vision of a billion people in VR has been quietly replaced with a new thesis: the face is the platform.

Production Is Doubling

In January 2026, Bloomberg reported that Meta and EssilorLuxottica were discussing a capacity increase from 10 million to 20 million units by the end of 2026, with potential to reach 30 million if demand held. Global expansion to the UK, France, Italy, and Canada had already been paused due to "unprecedented demand and limited inventory" in the US market. Waitlists were extending into the second half of 2026.

Twenty million annual units would match Apple Watch's peak annual volume. At that scale, Meta's installed base would cross 30 million cumulative devices by the end of 2027. For context, the iPhone crossed roughly 15 million cumulative units in early 2009, about two years after launch, and that is when third-party app development began to accelerate. If production hits 20 million this year, Meta's glasses could reach a comparable installed base in 2026.

An installed base alone does not guarantee a platform succeeds. But when 15+ million people are wearing cameras and microphones on their faces with an always-on AI assistant, the question shifts from "is this a real product?" to "what can I build for it?"

Competition Is Coming Late

Every major consumer electronics company has noticed the sales numbers. Google partnered with Gentle Monster and Warby Parker for Gemini-powered smart glasses. Samsung is planning a Meta Ray-Ban competitor powered by Google's Android XR for 2026. Apple reportedly moved engineers from its shelved Vision Air project to prioritize smart glasses development with a possible 2027 release. XReal raised $100 million at a $1 billion+ valuation and partnered with Google for Android XR integration.

None of these competitors have shipped a mass-market product yet. Meta has a 9-million-unit head start and, with EssilorLuxottica, the only proven supply chain partnership that can stamp out millions of optically correct, fashion-forward frames with embedded electronics. Replicating that relationship is not a software problem. It requires years of co-engineering between chip designers and lens grinders.

Strongest Counterargument

These glasses sell because they are Ray-Bans, not because they are smart.

A standard pair of Ray-Ban Wayfarers costs $163-$213. The Meta version starts at $299. Strip away the camera, speakers, microphone, and Meta AI integration, and most buyers would probably still pay close to $200 for the frame. The effective "smart" premium is roughly $100, which suggests the technology is a feature addition to existing eyewear, not a standalone computing platform.

Google Glass proved this exact failure mode. Without an iconic frame brand providing distribution and social camouflage, standalone smart glasses could not find a market. The "Glasshole" stigma killed the product faster than any technical limitation. Meta solved the social acceptance problem by hiding the computer inside a Wayfarer. That is a distribution strategy, not a platform strategy.

If this argument is correct, Meta is building an accessory business on top of EssilorLuxottica's brand equity, not a new computing platform. Accessories are good businesses. But they are not worth $83.5 billion in cumulative losses. The bull case requires the glasses to eventually become a developer platform with third-party apps, revenue-generating services, and lock-in that justifies the investment. As of March 2026, the only "apps" on the glasses are Meta AI and a camera. That is not a platform.

Limitations

EssilorLuxottica reports combined Ray-Ban and Oakley Meta sales. No breakout exists by model (Gen 1 vs. Gen 2) or by frame style (Wayfarer vs. Headliner vs. Skyler), so the 7 million figure includes all variants without revealing which configurations drive demand.

The 70/30 wearables-to-VR expense split is forward-looking guidance for 2026, not a confirmed historical ratio. Applying it retroactively to 2025 spending produces an approximate, not exact, subsidy-per-pair figure. Actual 2025 allocation may have been closer to 50/50 or 60/40, which would lower the per-pair subsidy to $1,213-$1,456.

Reality Labs revenue includes Quest hardware, software, and accessories alongside glasses revenue. Since Meta does not disclose glasses-specific revenue, the subsidy calculation treats all RL revenue as a single pool. If glasses generate meaningful revenue (say, $500M of the $2.21B), the net burn attributable to wearables would be lower.

Production capacity discussions (20-30 million units) reflect manufacturing intent, not committed orders. EssilorLuxottica and Meta have not finalized the expansion. Demand forecasting for a two-year-old product category carries substantial uncertainty.

Apple Watch annual sales figures are analyst estimates from firms including IDC and Canalys. Apple has never disclosed Watch unit sales. Comparisons to iPhone are more reliable because Apple reported unit volumes through 2018.

Usage data for Ray-Ban Meta glasses, the metric that matters most for a platform play, is not publicly available. Seven million pairs sold tells you nothing about whether those glasses are on faces or in drawers.

The Bottom Line

Meta is spending $1,698 for every $299 pair of glasses it ships. That math is terrible, and it does not matter yet. Platform bets are judged by adoption curves, not unit economics, and Meta's adoption curve is outpacing the iPhone's at the same stage. The 70/30 pivot from VR to wearables signals that even Meta now believes the face, not the headset, is where computing goes next. Whether 7 million buyers purchased a computer or just liked the sunglasses will determine if this is the early iPhone or the next Google Glass. At $83.5 billion and counting, Meta is betting everything that people are buying the future, not just the frames.