Subtitles on the lens, or a voice at your ear.

The G2 turned translation into subtitles

The G2 is the sequel to the most disciplined product in smart eyewear. Like the G1 before it, it refuses a camera and refuses speakers on purpose, and puts a quiet text display in the lens. What changed is the headline: this generation leads with translation— live captions of the language being spoken around you, drawn where your eye already is, in the dozens of languages the maker’s own list runs to.

We think that framing is correct, and we say so as a competitor: translation is the feature that makes strangers stop and ask where you got your glasses. The G2 validated it as the reason to wear a computer on your face. The open question — and the entire substance of this guide — is the delivery channel. Because subtitles are one way to receive a language. They are not the only way, and for a conversation they may not be the best one.

Reading a conversation vs. having one

Strip away the branding and there are exactly two designs for glasses-based translation in 2026, and they fork on a single question: does the translation arrive as text or as sound? Everything downstream follows.

• Where your eyes go.Captions put your gaze on floating text while a person is talking to you. In-ear audio keeps your eyes on their face. Eye contact is most of what makes a cross-language exchange feel like a conversation instead of a transaction — this is the audio design’s strongest card, and we’d argue it’s decisive for face-to-face use.
• Who gets to understand.A caption is private by physics: only the wearer sees it. That’s elegant for a lecture and lopsided for a dialogue — the other person has no channel back unless you hand over a phone. Open-ear audio played aloud is the opposite: both sides hear the translation, so the exchange runs two-way without passing hardware around.
• Where each one breaks. Honest symmetry: a loud market drowns open-ear audio, and text does not care how loud the room is. A silent meeting is where captions win outright — audio has no polite volume there. Pick by the rooms you actually stand in.
• What else the frame can be.Skipping the optical engine frees the weight and cost budget. That’s why an audio-first frame can carry an 8 MP camera and land at $149, while the caption tier sits in the high hundreds — the display is the most expensive component in smart eyewear, and you pay for it whether or not captions are your daily habit.

What the audio-first alternative covers, from $149

The OXIVUE line answers the same travel-and-conversation brief with sound instead of a screen:

• Two-way spoken translation. The M08 ($149), M02C ($159), and M08C/M08A ($169) cover 40+ languages; the M14 wrap shield ($199) covers 139 through the same free HeyCyan app. Both directions of the conversation, heard aloud. How the pipeline actually works, delay and all: our explainer.
• Open-ear directional audio. Speakers aim at your ear and leave the canal open — you hear the translation and the street at the same time. The M14 adds two noise-cancelling microphones plus a bone-conduction mic for wind and motion.
• A camera, because the weight budget allows one.Every frame carries an 8 MP Sony sensor — hidden at the bridge on the M08 family, stabilised 1080p video at 30 fps on the M14. The caption tier omits this by philosophy; here it’s the travel companion feature translation pairs best with.
• Ordinary-glasses weight.41–42 g across the M08 family, 52 g for the M14’s wrap format. Swappable photochromic, sun, or clear lenses on the M08 family; the M02C adds an IP65 rating, the strongest in our line.
• No subscription, no platform account.The app is free, translation carries no recurring fee, and nothing routes through a social platform’s cloud.

What the caption premium honestly buys

We don’t publish competitor spec tables — specs and prices move every season and a stale table helps nobody. The structural differences are what hold, and they are real:

• Silence. Captions work where audio is socially impossible — a meeting, a lecture, a quiet train. If most of your translation happens in rooms where you cannot make a sound, the display tier is the correct tool and no audio frame substitutes for it.
• Noise immunity. Text is unreadable by no crowd. In sustained high noise, captions beat any open-ear speaker, ours included.
• Zero etiquette load. No camera and no speakers means nobody around you wonders about recording or overhears your assistant. It is the least socially intrusive smart frame that exists, and that discipline is precisely what you are paying the high hundreds for — along with the optical engineering, and before the optional ring controller the ecosystem nudges you toward.

The honest limits of the audio-first design

Symmetry cuts both ways. OXIVUE frames have no display and no HUD — we cannot show you a caption, a teleprompter, or a notification, ever, at any price. In a genuinely loud room our open-ear translation degrades where a caption would not. We don’t offer prescription lenses — the working answer is contacts, and here’s the honest Rx state of play. Translation needs a phone data connection, same as the caption tier. And our 30-day returns are real but the return shipping is on you — we’d rather say that here than surprise you at the mailbox.

Who should buy which

Pay the caption premium if…

• Your translation happens where silence is mandatory — meetings, lectures, courtrooms.
• You mostly need to understand, not to be understood back.
• A camera on your face is a professional or personal non-starter, and budget in the high hundreds is comfortable.

Take the audio-first frame if…

• Your translation is face-to-face — markets, taxis, in-laws — and both sides need to follow it. See our travel guide for that life.
• You want the category’s marquee feature from $149 instead of the high hundreds, with 40+ languages on the M08 family or 139 on the M14.
• You want a camera in the same frame — the moment after the conversation is usually worth keeping.

Bottom line

The G2 proved translation is the reason smart glasses exist; its captions are the right answer for silent rooms and one-way understanding. But a conversation is two people, and sound is the only channel both of them share. The M08 at $149 speaks 40+ languages both ways with eye contact intact, and the M14 at $199 stretches that to 139 — no subscription, no account, 30-day returns, 2-year warranty. For the wider display-versus-camera philosophy this fork belongs to, read the G1 comparison; for every spec side by side, the comparison page; and if you’re budgeting, the under-$200 buyer’s guide.

Frequently asked questions