Halliday Launches G2 Meeting-Focused AI Glasses
Halliday introduced its G2 smart glasses on July 21, opening preorders for a $599 camera-free wearable expected to ship in September. The second-generation model uses dual monochrome waveguide displays and centers its software on Meeting Flow, which is designed to show live captions, translations, summaries, decisions, and follow-up prompts during workplace conversations.
Halliday introduced the G2 on July 21 as its second-generation AI glasses, with preorders opening the same day and shipments targeted for September 2026. The camera-free wearable is expected to retail for $599; a $10 preorder deposit is advertised with a $100 discount on the final price.
A new display and a narrower use case
Halliday's first-party announcement describes the G2 as display-equipped AI glasses built for meetings. The hardware replaces the original model's small DigiWindow display with a waveguide in each lens. iPhone in Canada reports dual monochrome micro-LED projectors with 600-by-300 resolution per eye and brightness of up to 1,600 nits, plus a four-microphone array and a target weight below 50 grams.
The Verge tested the new glasses before launch and found the dual display easier to see than the first-generation design in an indoor setting. Its reviewer did not test visibility in direct sunlight or establish long-term reliability. Halliday told the publication to expect about 12 hours of typical use, a vendor estimate that still needs independent battery testing.
Meeting Flow is the main software bet
The product's central feature is Meeting Flow. Halliday says the system can surface live captions, translations, short summaries, and contextual prompts while a conversation is happening. Thread Tracker is meant to show how topics develop, Decision Confirmation to identify conclusions, and Commitment Check to flag whether an action has an owner and deadline. Post-meeting transcripts, summaries, and open items are presented in the companion app.
The glasses also support voice control, calls, notifications, teleprompter-style notes, and translation in more than 45 languages, according to the two launch reports. With no camera, the device cannot use visual context in the way camera-equipped AI glasses can. Halliday frames that omission as a workplace-privacy choice and a way to reduce weight and power use.
What still needs proof
A camera-free design removes video capture, but it does not settle the privacy questions created by microphones, transcription, and cloud-connected AI. Buyers and enterprise teams still need clear answers about consent, retention, access controls, and whether audio or transcripts train future systems. The retrieved launch material does not provide those implementation details.
Meeting Flow also combines several error-prone tasks: streaming speech recognition, speaker tracking, translation, summarization, and extraction of decisions and commitments. A missed negation or wrongly assigned action item could be more disruptive than no prompt at all. The Verge saw only a brief demonstration and explicitly left the usefulness of the AI features unresolved. The September release will provide the first broader test of whether the product can perform reliably in noisy, multilingual, and high-stakes meetings.
Key Points
- 1Halliday opened G2 preorders on July 21 at an expected $599 retail price, with shipments targeted for September 2026.
- 2The camera-free glasses use dual monochrome waveguide displays and position Meeting Flow as their central workplace feature.
- 3The product's practical value still depends on live transcription accuracy, latency, consent controls, and reliable extraction of decisions and action items.
Scoring Rationale
The G2 is a notable wearable-AI product because it combines live transcription, translation, summaries, and contextual prompts with a heads-up display. Its practitioner impact remains limited by undisclosed model and data-governance details and by the need to validate accuracy, latency, privacy, and usability outside short demonstrations.
Sources
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
