
Meta's leaked memo reveals AI pendant, supersensing glasses, and enterprise wearables strategy
Quick Answer
Meta plans to internally test an AI pendant by spring 2027 and expand its smart glasses lineup with 'supersensing' models.
Quick Take
Meta plans to internally test an AI pendant by spring 2027 and expand its smart glasses lineup with 'supersensing' models. The strategy aims to curb losses in Reality Labs while targeting corporate customers with a new enterprise offering called 'Wearables for Work.'
Key Points
- AI pendant testing set for spring 2027; could feature a camera.
- Meta's smart glasses sales target is 10 million devices by late 2026.
- New subscription model for Meta AI offers tiers priced at $7.99 and $19.99.
- Daily usage of AI-powered smart glasses has tripled year-over-year.
- Meta aims for 6.8 million monthly active wearable users by year-end.
📖 Reader Mode
~3 min readMeta wants to test an AI companion device internally next year and broaden its range of AI-powered smart glasses.
That's according to an internal memo from Alex Himel, Meta's Vice President of Wearables, which The Information was able to review. The strategy aims to stop the massive losses at the Reality Labs hardware division while pushing adoption of Meta's AI models and products.
Himel lays out three pillars in the memo: a new AI pendant, an expanded smart glasses lineup, and an enterprise offering called "Wearables for Work." The last one targets corporate customers who'd be more willing to pay for devices with industry-specific features.
The devices will run on Meta's latest AI model, Muse Spark, and an unreleased AI agent called "Hatch."
AI pendant headed for internal testing by spring 2027
The memo doesn't include specs for the planned AI pendant, according to The Information, but the device could have a camera. Internal testing—known as "dogfooding" in industry jargon—is set to start in spring 2027. Meta already acquired the AI pendant startup Limitless last year.
Beyond the pendant, Meta is expanding its glasses lineup to include so-called "supersensing" models. Super Sensing keeps cameras and sensors running for hours at a time, letting the AI assistant track what happened throughout the day. Meta AI could notice when you've forgotten your keys or remind you to pick up an ingredient for dinner.
So far, Meta's glasses lineup has been limited to Ray-Ban Meta and Oakley Meta through a partnership with EssilorLuxottica. More brands and styles are coming to reach a wider audience and improve gross margins. EssilorLuxottica reported in February that more than 7 million Meta-powered smart glasses sold in 2025.
Daily use of AI-powered smart glasses has tripled year over year, according to Zuckerberg, who calls it "one of the fastest-growing categories of consumer electronics ever." Himel has set a target of ten million wearable devices sold in the second half of 2026.
Software subscriptions to offset hardware losses
Himel wants to hit 6.8 million monthly active wearable users by year's end. New products and sales in more countries are the levers to get there, according to the memo.
Meta also aims to make money on software. This week, the company rolled out a two-tier subscription for its AI chatbot Meta AI, available through Facebook, Instagram, WhatsApp, and soon on its smart glasses.
Meta One Plus ($7.99) and Meta One Premium ($19.99) follow the same playbook as OpenAI and Google: users pay for more compute, longer model reasoning, and more image and video generation. "All of our glasses are designed to easily update to use our newest AI models and features," Zuckerberg said. Meta is also building a developer platform where third parties can upload apps for the wearables.
Meta isn't alone in its AI hardware ambitions. OpenAI acquired io Products, co-founded by ex-Apple designer Jony Ive, for $6.5 billion and reportedly has more than 200 people working on a family of AI devices, including a smart speaker in the $200 to $300 range and possibly a smartphone. Google has announced smart glasses in partnership with Samsung and Qualcomm for this fall.
— Originally published at the-decoder.com
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.

