The AR/VR category enters 2026 with three distinct sub-segments forming, each with its own battery-sourcing reality. For pouch-cell suppliers, these segments are not equally attractive — and the wrong bet could leave a factory stranded on cell sizes that nobody orders in 2027.
Three segments, three battery worlds
1. Standalone VR (tethered experience, heavy)
This is the mature segment. Devices are roughly 400–650 g, use 3,500–5,500 mAh cells, and the industrial-design pressure on the battery has plateaued. Customers in this segment buy in large volumes and ask for long-cycle-life (800+) pouch cells at conservative cell chemistries — 4.35–4.40 V platforms are the norm.
Supply is not tight here. Multiple Chinese and Korean suppliers ship volumes in the tens of millions a year, and pricing has been trending down 6–9% year-over-year since 2023.
2. AR glasses (all-day, lightweight)
The fastest-growing segment. Form factors are converging around 40–80 g headsets with 150–500 mAh ultra-thin cells. This is where 4.48 V high-voltage LCO is displacing conventional LCO, and where suppliers who can hold < 1.0 mm cell thickness tolerance have a real moat.
Demand signals are bullish but choppy. Several consumer-electronics OEMs are taking multi-million-unit bets on new generations launching in late 2026 and 2027. If one of those launches succeeds, the industry will be short on 0.6–0.9 mm pouch cell capacity by the second half of 2026.
3. Smart glasses (display-less, audio-first)
The volume leader in unit terms but not in cell consumption. A sensor-only or audio-only smart glass uses a 30–80 mAh coin or micro-pouch cell; total lithium content is under a gram. These devices ship in tens of millions a year at low cell prices, but the margin-per-cell is thin. It’s a segment to serve if you already have a coin-cell line; not one to build a factory around.
What suppliers are hearing from customers
Conversations with battery buyers in this category in Q1 2026 landed on four consistent themes:
- Density trumps cycle life for AR glasses — customers accept 400-cycle ratings if runtime increases.
- Thermal envelope is being specified more aggressively. Skin-contact limits are now in RFQs, not discovered during EVT.
- Swelling is auditable. Several tier-1 customers now require swelling data at multiple aging points, not just an end-of-life number.
- Dual sourcing is mandatory. After 2023’s supply disruptions, no serious AR program is single-sourcing.
The bottleneck nobody’s talking about
The constraint in AR battery supply is not cell manufacturing capacity — it’s electrode coating lines capable of sub-35 µm coatings at tight uniformity. That equipment is specialised, mostly built by Japanese and Korean OEMs, and deliveries in 2026 are being allocated quarter by quarter. Any battery supplier making commitments for ultra-thin volumes should check the coating-line book before promising lead times.
What to bet on
For a supplier positioning for the next 24 months:
- Invest in 4.48 V platform qualification now. Customers will not wait for you to catch up in late 2026.
- Build excess pilot-line capacity for 0.6–1.0 mm cells. AR design cycles are fast; the team that can turn a new cell geometry in 4 weeks wins.
- Don’t chase VR volumes. Margins are compressed and the segment is already served.
- Offer BMS-integrated packs, not bare cells. AR customers increasingly buy fully-integrated solutions to close faster.