South Korean startup LetinAR ships optical modules for AI glasses

LetinAR secures $18.5 million for optical module scale-up

LetinAR, a South Korean startup founded in 2016, raised $18.5 million from Korea Development Bank and Lotte Ventures ahead of a planned 2027 IPO in South Korea. The capital will fund manufacturing scale-up as AI glasses move from early adoption to mass production. The company's total funding now stands at $41.7 million (company-reported).

LetinAR does not make glasses. It manufactures the optical module, the thumbnail-sized lens component that projects images into the wearer's field of vision. This module determines whether a pair of smart glasses feels like laboratory hardware or something usable for work and daily life. The engineering constraints are severe: the lens must be thin, light, power-efficient, and produce a sharp, bright image, all within a form factor that fits inside a normal-looking frame.

The company has already shipped modules to customers including Japan's NTT QONOQ Devices and Dynabook (formerly Toshiba Client Solutions). Its most visible customer is Aegis Rider, a Swiss deeptech firm building an AI-powered AR motorcycle helmet that displays real-time navigation and safety alerts anchored to the road itself. Aegis Rider targets European markets in 2026.

Optical modules are the bottleneck the industry cannot solve alone

Global AI glasses shipments surged to 8.7 million units in 2025, up 300% from the prior year, with projections exceeding 15 million units in 2026 (per Omdia). Major players have entered or are entering the market: Meta has sold AI-enabled Ray-Ban glasses since 2023; Google is building Android XR; Apple is expected to announce a product; Samsung is reportedly unveiling co-designed glasses this July; and Chinese manufacturers including Huawei, Alibaba, and Xiaomi are all competing.

Despite this momentum, no glasses maker has solved the optical module problem at scale. LetinAR's technology, called PinTILT, arranges microscopic optical elements inside a lens to direct light precisely into the user's eye rather than scattering it across the full lens surface. The company claims this approach delivers brighter images in thinner, lighter form factors while using less power than existing waveguide-based designs, which split light broadly across the lens and waste much of it before it reaches the eye.

The alternative, mirror-based birdbath optics, delivers light more efficiently but is too bulky to fit inside a normal-looking frame. LetinAR positions PinTILT as the middle path, but the company has not published independent optical benchmarks. Competitors including WaveOptics, DigiLens, and Lumus are pursuing similar module designs and also lack published third-party validation of their efficiency claims.

Validate supplier claims before committing

LetinAR's customer roster and manufacturing experience are real signals of progress. The Aegis Rider use case, anchoring information to the physical world rather than the visor, is a meaningful application that requires precise optical calibration. Shipping at scale to established manufacturers like Dynabook adds credibility.

However, the optical module supply space remains unsettled. No independent benchmark has reproduced LetinAR's efficiency claims or compared PinTILT directly against waveguide or birdbath alternatives under controlled conditions. Vendor-published performance data, absent external validation, is insufficient for locked-in contracts. Product teams selecting optical suppliers should request access to reproducible test data and consider multi-vendor roadmaps until the performance tradeoffs between approaches are independently confirmed.