AURAI
KollektionenMarkenAkademieGlossarKI-BeraterVergleichen

AURAI

Defining the future of wearable vision. We analyze, score, and recommend the world's most advanced smartglasses.

Platform

CatalogBrandsCompareAcademyHow We Rate

Collections

AudioPhotographyNavigationWork

Transparency

AURAI is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com. Some links on this site may earn us a small commission at no additional cost to you.

© 2026 AURAI Smartglasses - A brand of Eisenlohr Consulting

|

UID: CHE-399.545.184

Privacy PolicyTerms of Service
    Home/Glossary
    AURAI Glossary

    Smart Glasses Specs, Decoded

    Every spec explained with AURAI's critical perspective. Learn how manufacturers measure each metric, how they game the numbers, and what really matters for your daily use.

    89 terms

    Brightness

    brightness_nits
    Display & OpticsUsed by 55 products

    Display brightness in nits (cd/m2).

    The Critical View

    Brightness specs in smart glasses are perhaps the most misleading of all display metrics. Problems include: (1) manufacturers report panel brightness before optical losses — a panel rated at 2,000 nits behind a waveguide with 40% optical efficiency delivers only ~800 nits to the eye; (2) "peak brightness" vs. "sustained brightness" vs. "typical brightness" are conflated — peak brightness may only be achievable in a small HDR highlight window or for a few seconds before thermal throttling; (3) nits measured at the panel do not account for ambient light washout — a 1,000-nit see-through display becomes nearly invisible in direct sunlight (~100,000 lux) because the human eye adapts to ambient brightness; (4) some manufacturers quote "perceived brightness" which is a subjective metric with no standard definition.

    How It's Measured

    Integer value in nits (cd/m2). AURAI stores the manufacturer's stated brightness figure, which typically represents peak panel brightness before optical losses. For accurate outdoor usability assessment, divide the stated value by 2–3x for birdbath optics or 3–5x for waveguide optics to estimate eye-level brightness.

    Real-World Feel

    Under 500 nits (at-eye): indoor use only — display washes out near windows. 500–1,500 nits (at-eye): usable indoors and in shade, struggles in direct sunlight. 1,500–4,000 nits (at-eye): visible in most outdoor conditions except direct sunlight glare. Above 4,000 nits (at-eye): truly sunlight-readable, currently rare. For context, a 3,000-nit panel behind a typical waveguide (35% efficiency) delivers roughly 1,050 nits to the eye — adequate for overcast outdoor use but not sunny conditions. Most consumer AR glasses in 2026 deliver 400–1,200 nits at-eye in practice.

    Deep dive →

    Diopter Adjustment

    diopter_adjustment
    Display & OpticsUsed by 97 products

    Built-in vision correction adjustment.

    The Critical View

    This is a significant usability factor that manufacturers often bury in footnotes. Without diopter adjustment, users who wear prescription glasses must either: (a) wear the smart glasses over their existing glasses (uncomfortable, often impossible due to fit), (b) purchase custom prescription lens inserts ($75–$200 additional cost and 1–3 week wait), or (c) wear contact lenses. Manufacturers market "prescription compatible" which may only mean option (b) exists as a third-party accessory, not that the device has built-in correction. The distinction matters enormously for the ~65% of adults who need vision correction. Some devices offer adjustable diopter dials (typically -5.0D to +3.0D range) while others have fixed focal distances that work for some prescriptions but not others.

    How It's Measured

    Boolean. TRUE if the product has any form of built-in, user-adjustable diopter correction (mechanical dial, slider, or electronic). Separate prescription insert accessories do NOT qualify — those are tracked via the prescription_lens_option field. The specific diopter range is stored in diopter_adjustment_range.

    Real-World Feel

    If TRUE with a range of -5.0D to +2.0D (common), the device serves roughly 80% of nearsighted users without additional accessories. Users with astigmatism correction needs are generally NOT served by simple diopter dials — they still need custom inserts. If FALSE, budget an additional $75–$200 and 1–3 weeks for prescription inserts, or plan to wear contact lenses. This field is a critical purchase decision factor for the majority of potential buyers.

    Display Optics Technology

    display_optics_tech
    Display & OpticsUsed by 86 products

    Optical relay method (Birdbath, Waveguide, Freeform Prism).

    The Critical View

    This is where the most aggressive marketing obfuscation occurs. "Waveguide" sounds premium but encompasses vastly different quality levels — from single-layer diffractive waveguides with severe rainbow artifacts and <20% optical efficiency to multi-layer reflective waveguides with >40% efficiency and minimal color artifacts. Manufacturers almost never disclose waveguide type (diffractive vs. reflective vs. holographic) or layer count, which determine actual image quality. Similarly, "Birdbath" is treated as inferior in marketing despite delivering consistently good image quality with less color distortion than cheap waveguides. The real trade-off is form factor: waveguides enable thinner, more glasses-like designs while birdbath optics require bulkier housings.

    How It's Measured

    Enum values: Birdbath (half-mirror combiner — simple, reliable, bulkier), Waveguide (light guided through glass substrate — thinner form factor, variable quality), Freeform Prism (shaped prism element — compact but limited FOV), None (no optical relay — direct view or non-display product). Verified through manufacturer specs, patent filings, and optical teardowns.

    Real-World Feel

    Birdbath optics typically achieve 50–80% of the panel's native image quality with even illumination and minimal artifacts, but create a distinctive "sunglasses" tint. Waveguide optics can look nearly transparent when the display is off (ideal for daily wear) but introduce color non-uniformity, reduced brightness (30–50% light loss is typical in diffractive designs), and eye-box sensitivity — move your eye slightly off-center and the image dims or disappears. Freeform prisms offer a good balance for monocular displays but constrain FOV to 15–25 degrees.

    Deep dive →

    Display Panel Technology

    display_panel_tech
    Display & OpticsUsed by 77 products

    Micro-display technology (Micro-OLED, LCoS, LCD, Micro-LED).

    The Critical View

    Manufacturers frequently conflate panel tech with optics tech. "Waveguide display" is a common marketing phrase that describes the optics, not the panel — the actual panel behind a waveguide is typically LCoS or Micro-LED. AURAI separates these into display_panel_tech and display_optics_tech to prevent this confusion. Watch for "OLED" claims that are actually referring to a companion device screen, not the eyepiece. Some brands also use "Sony Micro-OLED" as a prestige signal without disclosing the specific panel model or generation, despite meaningful differences between Sony ECX series revisions.

    How It's Measured

    Sourced from manufacturer datasheets, FCC filings, and teardown analyses. Enum values: Micro-OLED (self-emissive, best contrast, most common in consumer AR displays), LCoS (reflective, common in waveguide-based enterprise AR), LCD (transmissive, cheapest but worst contrast), Micro-LED (emerging, highest potential brightness). The "Waveguide" and "Birdbath" enum values are legacy data artifacts and should be migrated to display_optics_tech where encountered.

    Real-World Feel

    Micro-OLED delivers the deepest blacks and best contrast ratios (typically >100,000:1), making it ideal for media consumption in dim environments. LCoS panels offer acceptable sharpness but struggle with contrast (typically 1,000–5,000:1) and exhibit noticeable rainbow artifacts. LCD panels are the budget option with worst black levels. Micro-LED promises superior brightness for outdoor use (>10,000 nits theoretical) but remains rare and expensive in 2026.

    Deep dive →

    Display Present

    display_present
    Display & OpticsUsed by 97 products

    Indicates if the device has a visual display.

    The Critical View

    Many audio-only smart glasses (e.g., Ray-Ban Meta, Bose Frames) are marketed alongside display-equipped devices, creating false equivalence in comparison tables. Some manufacturers blur the line by calling LED notification lights or rudimentary status indicators a "display." AURAI enforces a strict definition: this field is TRUE only when the device projects imagery the user views through the optics — not indicator LEDs, not companion-app screen mirroring, and not camera viewfinders.

    How It's Measured

    Binary field. Verified against manufacturer spec sheets and hands-on reviews. Edge cases: products with a tiny monochrome OLED status bar (e.g., some Vuzix enterprise models) are marked TRUE because the user reads information through the optics.

    Real-World Feel

    If FALSE, the product is fundamentally an audio/camera wearable that happens to be shaped like glasses. Nearly all comparison shopping (FOV, resolution, brightness) is irrelevant for display_present=FALSE products, so this field gates whether other display specs are meaningful.

    Electrochromic Dimming

    electrochromic_dimming
    Display & OpticsUsed by 97 products

    Electronic lens tinting (darken/lighten on demand).

    The Critical View

    Electrochromic dimming is a premium feature currently found on only a handful of products (notably XREAL Air 2 Pro, Viture Pro XR, and select others). Manufacturers sometimes conflate this with photochromic lenses (which darken passively in UV light, like Transitions lenses) — the key distinction is that electrochromic is electronically controlled (instant, user-toggled, works indoors) while photochromic is passive (slow transition, UV-dependent, does not work behind car windshields that block UV). Some products offer only 2–3 fixed tint levels rather than continuous adjustment, which manufacturers may not disclose clearly.

    How It's Measured

    Boolean. TRUE only for electronically controlled dimming (voltage-activated). Passive photochromic (UV-reactive) lenses are NOT electrochromic and should be tracked in lens_treatment instead. The number of tint levels and dimming range are not currently captured as separate fields.

    Real-World Feel

    Electrochromic dimming transforms a product from "indoor only" to "indoor/outdoor versatile." Without it, see-through AR displays wash out in bright environments, and media-consumption glasses require physically swapping lens covers or clip-on shades. With electrochromic dimming, users can toggle from ~0% tint (transparent, for indoor AR overlays) to ~30–60% tint (comfortable outdoor use) electrically. This is particularly valuable for products that double as everyday sunglasses. The technology adds $50–$100 to manufacturing cost, which is why it remains a premium differentiator.

    Deep dive →

    Field of View (Diagonal)

    field_of_view_diagonal
    Display & OpticsUsed by 56 products

    Diagonal field of view in degrees.

    The Critical View

    FOV is the single most manipulated spec in smart glasses marketing. Manufacturers game it in at least four ways: (1) reporting diagonal FOV to inflate the number vs. horizontal or vertical, which matter more for actual usability; (2) measuring at the absolute optical edge where brightness and sharpness have fallen off significantly, rather than at the usable or "90% brightness" boundary; (3) quoting "equivalent screen size" (e.g., "equivalent to a 140-inch screen at 3 meters") which is mathematically derived from FOV but sounds more impressive and hides the actual degree measurement; (4) conflating monocular and binocular FOV in stereo overlap configurations. There is no industry-standard measurement protocol for smart glasses FOV, unlike the ISO 13007 standard used for head-mounted displays in simulation.

    How It's Measured

    Numeric value in degrees. Diagonal measurement. AURAI attempts to normalize to the manufacturer's stated diagonal FOV but flags discrepancies when hands-on reviews (e.g., Karl Guttag's blog, Road to VR measurements) diverge by more than 15% from claimed specs. Horizontal FOV can be roughly estimated as diagonal * cos(atan(h/w)) using the display aspect ratio.

    Real-World Feel

    Under 20 degrees: feels like a tiny floating postage stamp — only suitable for quick-glance notifications. 20–35 degrees: equivalent to a small tablet held at arm's length — usable for basic AR overlays and short video clips. 35–50 degrees: the sweet spot for "virtual monitor" products — feels like a 60–100 inch screen at comfortable viewing distance. 50–70 degrees: genuinely immersive for video, starts to enable spatial computing. Above 70 degrees: rare in glasses form factor, typically headset territory. Most consumer AR glasses in 2026 claim 45–52 degrees diagonal.

    Deep dive →

    Lens Light Transmission

    lens_light_transmission_pct
    Display & OpticsUsed by 39 products

    Visible Light Transmission percentage (how transparent the lens is).

    The Critical View

    VLT (Visible Light Transmission) is an objective, measurable spec — but manufacturers rarely disclose it, preferring subjective terms like "lightly tinted" or "sunglasses-grade." When they do provide VLT, it may be measured at a single wavelength (often 550nm, peak eye sensitivity) rather than across the full visible spectrum, which can overstate transparency for lenses with selective color filtering. For electrochromic devices, manufacturers may quote only the lightest setting's VLT without disclosing the darkest, or vice versa. For see-through AR displays, VLT interacts with display brightness to determine outdoor usability — high VLT (transparent lens) lets in more ambient light, requiring brighter displays to maintain contrast.

    How It's Measured

    Numeric value, 0–100 percentage. 100% = perfectly transparent, 0% = fully opaque. Measured per ISO 12312-1 for sunglasses or per EN 166 for safety eyewear, though smart glasses manufacturers rarely cite which standard they used (if any). For products with adjustable tint, we attempt to store the lightest (most transparent) setting.

    Real-World Feel

    80–100% VLT: essentially clear lenses, suitable for indoor use and nighttime. 40–80% VLT: light tint, comfortable for overcast or indoor-to-shade transitions; the "daily wear" sweet spot for smart glasses. 18–40% VLT: moderate sunglasses territory, comfortable in daylight, too dark for indoor use. 8–18% VLT: dark sunglasses, strong sun protection, suitable for bright outdoor conditions. Below 8% VLT: extreme conditions only (snow, water glare). Most display-equipped smart glasses land at 30–50% VLT, balancing see-through transparency with enough tint to improve display contrast.

    Lens Treatment

    lens_treatment
    Display & OpticsUsed by 54 products

    Lens coating/treatment type (e.g., polarized, photochromic, anti-reflective).

    The Critical View

    Lens treatments vary enormously and manufacturers use inconsistent terminology. "Anti-glare" may refer to a basic AR coating or to polarization — very different technologies. "Blue light filtering" has minimal scientific evidence for eye health benefits but is marketed aggressively. Polarized lenses genuinely reduce glare from reflective surfaces but can make LCD screens (phones, car dashboards, ATMs) unreadable at certain angles — a significant drawback for smart glasses that pair with phones. Photochromic treatments (auto-darkening in UV) are useful but typically take 30–60 seconds to transition and may not fully clear indoors. Premium treatments like Oakley Prizm use selective wavelength filtering to enhance contrast for specific activities, which is legitimately different from simple tinting.

    How It's Measured

    Free text field. Common values include: "Clear", "Tinted", "Polarized", "Photochromic", "Anti-reflective (AR coated)", "Blue light filtering", and brand-specific names like "Prizm Polarized 24k" or "Transitions XTRActive." No standardized vocabulary exists across the industry; AURAI normalizes where possible but preserves manufacturer terminology for brand-specific treatments.

    Real-World Feel

    For outdoor use, polarized lenses are the gold standard for comfort but check phone screen compatibility. For indoor/outdoor versatility, photochromic is convenient but slow to transition. For display-equipped glasses, anti-reflective coatings on the inner surface are critical to prevent the display image from bouncing back as a distracting ghost reflection. Clear lenses suit indoor-only or low-light use. The lens treatment choice significantly affects both visual comfort and the perceived quality of any embedded display.

    Refresh Rate

    refresh_rate_hz
    Display & OpticsUsed by 54 products

    Display refresh rate in Hertz.

    The Critical View

    Refresh rate in smart glasses is less straightforward than in monitors or phones. Manufacturers commonly spec the panel's maximum refresh rate without disclosing that the actual rendering pipeline — especially on tethered devices receiving video over USB-C — is limited to 60 Hz or even 30 Hz by the source device or cable bandwidth. A "120 Hz display" driven by a 60 Hz video signal delivers zero smoothness benefit. Additionally, some devices have variable refresh rate capabilities but market only the peak number. For AR overlay devices, the more critical metric is motion-to-photon latency (the delay between head movement and display update), which refresh rate alone does not capture.

    How It's Measured

    Integer value in Hz. Represents the panel's native maximum refresh rate as stated by the manufacturer. AURAI does not currently verify whether the full rendering pipeline supports the claimed rate, though we flag known discrepancies in editorial content.

    Real-World Feel

    60 Hz: acceptable baseline for video consumption and static AR overlays, but noticeable judder during head motion. 90 Hz: comfortable for AR/MR with moderate head tracking — reduces motion sickness risk. 120 Hz: smooth for fast-paced content and responsive head-tracked experiences. For media-consumption glasses without head tracking (e.g., XREAL Air, Rokid Max), refresh rate above 60 Hz primarily benefits gaming via compatible hosts. Most users cannot distinguish 90 Hz from 120 Hz in non-gaming AR use cases.

    Resolution per Eye (Height)

    resolution_per_eye_h
    Display & OpticsUsed by 57 products

    Vertical resolution per eye in pixels.

    The Critical View

    The same manufacturer inflation tactics apply as with horizontal resolution. Additionally, some products have non-standard aspect ratios (e.g., 16:9 panel in a 16:5 visible window) where the panel resolution exceeds the optically usable resolution — the edges are physically clipped by the optics. Vertical resolution is particularly important for text-heavy AR use cases like teleprompting and notification overlays, where vertical pixel density determines how many readable lines fit in the display.

    How It's Measured

    Integer value representing vertical pixels for a single eye. Should reflect the optically usable vertical resolution, not just the panel spec, though in practice most sources report panel resolution.

    Real-World Feel

    For wearable displays primarily used as virtual screens (movie watching, desktop mirroring), a vertical resolution of at least 1080 pixels is the practical minimum for comfortable extended use. Below 720 vertical pixels, text rendering quality drops noticeably. Notification-only AR overlays (e.g., INMO Air, Focals by North) may use as few as 480 vertical pixels, which is adequate for short text snippets but unsuitable for productivity.

    Deep dive →

    Resolution per Eye (Width)

    resolution_per_eye_w
    Display & OpticsUsed by 57 products

    Horizontal resolution per eye in pixels.

    The Critical View

    Manufacturers exploit resolution reporting in three primary ways. First, binocular products sometimes report combined (both-eye) resolution — a "3840x1080" spec might actually be 1920x1080 per eye. Second, panel resolution does not equal perceived resolution: optical losses, especially in waveguide systems, can degrade effective resolution by 20–40%, making a 1920-pixel panel look closer to 1200 pixels in practice. Third, resolution is meaningless without FOV context — 1920 pixels spread across 46 degrees is sharp (42 PPD), but the same resolution across 90 degrees is noticeably pixelated (21 PPD). Always compute pixels-per-degree (PPD) for meaningful comparison.

    How It's Measured

    Integer value representing horizontal pixels for a single eye. Sourced from panel manufacturer datasheets where available, otherwise from product specs. For binocular products using a single wide panel (e.g., some VR-style viewers), this should be half the total horizontal resolution.

    Real-World Feel

    Under 30 PPD (pixels per degree): visible screen-door effect and pixelation on text. 30–40 PPD: acceptable for video, readable large text. 40–60 PPD: sharp for most content, comfortable for extended reading. Above 60 PPD: retina-grade, indistinguishable from a high-quality monitor. Most consumer AR glasses in 2026 land in the 30–45 PPD range. For a typical 46-degree FOV display, 1920 horizontal pixels yields roughly 42 PPD — adequate but not exceptional.

    Deep dive →

    Supports 3d

    supports_3d
    Display & OpticsUsed by 97 products

    Stereoscopic 3D content support.

    The Critical View

    "3D support" is frequently overstated. True stereoscopic 3D requires independent left/right eye images with correct parallax — this demands either two separate displays (one per eye) or a single display with fast temporal multiplexing and synchronized shutter optics. Many binocular devices that clone the same image to both eyes are not stereoscopic even though they have two screens. Additionally, "3D support" may mean the hardware is physically capable but the software ecosystem has minimal 3D content available. Some manufacturers claim 3D capability based on side-by-side (SBS) video playback support, which is technically 3D but a low bar — any dual-display device can do this. True volumetric or depth-correct AR overlay is a much higher standard.

    How It's Measured

    Boolean. TRUE if the device supports independent per-eye rendering for stereoscopic depth. Verified against product documentation and hands-on testing. Devices that only display 2D content identically to both eyes are marked FALSE even if they are binocular.

    Real-World Feel

    For media consumption, 3D support enables immersive movie viewing (SBS 3D content libraries are growing) and spatial video playback (Apple Spatial Video, Meta 3D photos). For productivity, 3D is largely irrelevant. For gaming and spatial computing, it is essential. Note that 3D content induces vergence-accommodation conflict at fixed focal distances, which can cause eye strain in sessions longer than 30–45 minutes for sensitive users.

    Audio Frequency Response (Max)

    audio_freq_response_max_hz
    Audio & CameraUsed by 43 products

    Highest reproducible audio frequency in Hz.

    The Critical View

    The upper frequency limit is less commonly gamed than the lower limit because small drivers can physically reach high frequencies more easily. However, manufacturers still sometimes claim 20,000 Hz (the theoretical limit of human hearing) when independent measurements show significant rolloff above 12,000-16,000 Hz. For open-ear speakers, high-frequency performance is more affected by the acoustic coupling between the driver and the ear — small changes in glasses fit can shift the effective upper limit by several thousand Hz. Most adults over 30 cannot hear above 15,000-16,000 Hz regardless, making claims above that range irrelevant for the majority of users.

    How It's Measured

    Integer in Hertz (Hz). The same measurement standard inconsistencies apply as with the minimum frequency: -3 dB vs. -10 dB vs. unspecified. Common claimed values range from 8,000 Hz to 20,000 Hz. Prefer independent review measurements when available.

    Real-World Feel

    Below 8,000 Hz: speech remains intelligible but music sounds muffled and dull — cymbals, string harmonics, and vocal sibilance are lost. 8,000-12,000 Hz: adequate for speech-centric use (calls, podcasts, voice assistants). 12,000-16,000 Hz: the practical range for enjoyable music listening on glasses speakers. Above 16,000 Hz: inaudible to most adults; claims beyond this are marketing rather than functional.

    Deep dive →

    Audio Frequency Response (Min)

    audio_freq_response_min_hz
    Audio & CameraUsed by 43 products

    Lowest reproducible audio frequency in Hz.

    The Critical View

    This spec is routinely overstated for smart glasses. A manufacturer claiming "20 Hz" bass response on a miniature open-ear speaker is technically measuring the point where output drops to inaudibility, not where useful bass exists. In hi-fi audio, frequency response is measured at -3 dB; in marketing, it's often measured at -10 dB or even -20 dB, dramatically inflating the range. For the tiny drivers in smart glasses temples (typically 8-12mm), physically meaningful bass below 100-150 Hz is extremely difficult to achieve regardless of what the spec sheet claims. Bone conduction drivers have even worse low-frequency performance, rarely delivering perceptible output below 200 Hz.

    How It's Measured

    Integer in Hertz (Hz). Ideally measured at -3 dB relative to the reference level across the frequency band, but manufacturers almost never disclose the measurement standard. When available, prefer independent review measurements over manufacturer claims. Common claimed values range from 20 Hz to 200 Hz, but real-world useful output for glasses-form-factor speakers rarely extends below 100-150 Hz.

    Real-World Feel

    Below 80 Hz: deep bass (kick drums, sub-bass in electronic music) — essentially nonexistent in any current smart glasses. 80-150 Hz: upper bass — achievable by the best open-ear speakers, gives music some warmth. 150-300 Hz: lower midrange — the realistic floor for bone conduction and most glasses speakers. A claimed 20 Hz floor on a glasses speaker is marketing fiction. Users seeking actual bass response should pair glasses with separate earbuds.

    Deep dive →

    Camera Present

    camera_present
    Audio & CameraUsed by 97 products

    Indicates if the device has a built-in camera.

    The Critical View

    Some manufacturers list "camera-ready" or "camera compatible" when the hardware slot exists but no sensor is installed — the device ships without a functional camera. Others bundle a detachable camera accessory and market the glasses as "with camera" even though the base SKU excludes it. Always verify whether the camera ships in-box and is operational out of the box, not just physically possible.

    How It's Measured

    Boolean. True means a camera sensor is physically present and functional in the shipping product. False means no camera hardware exists. Do not set true for devices that merely support an optional camera accessory sold separately.

    Real-World Feel

    A true value here is the minimum bar for photo/video capture and visual AI features (scene recognition, live translation overlays). For privacy-conscious buyers, camera_present = true also means the device may face restrictions in certain venues, workplaces, and jurisdictions. Some users specifically seek camera-free models to avoid social friction.

    Camera Resolution

    camera_resolution_mp
    Audio & CameraUsed by 45 products

    Camera sensor resolution in megapixels.

    The Critical View

    Megapixels are one of the most misleading specs in consumer electronics. In smart glasses, the sensor is typically 1/4" to 1/3" — dramatically smaller than a smartphone sensor (1/1.3" or larger). A 12 MP glasses camera does not produce results anywhere close to a 12 MP smartphone camera. Manufacturers almost never disclose sensor size, pixel pitch, or lens aperture for eyewear cameras, making MP the only published number — and therefore the most abused. Higher MP on a tiny sensor actually degrades low-light performance because each photosite captures less light.

    How It's Measured

    Stated in megapixels (millions of pixels). Derived from sensor's effective pixel count (e.g., 4000x3000 = 12 MP). Watch for interpolated/upscaled resolution claims — some devices capture at a lower native resolution and software-upscale to a higher advertised MP count. Always prefer native sensor resolution.

    Real-World Feel

    Sub-5 MP: adequate for basic documentation and QR/barcode scanning but visibly soft when viewed on a phone screen. 8-12 MP: usable for social media sharing in good lighting; struggles indoors and at night. 12-16 MP: current high-end for smart glasses (e.g., Ray-Ban Meta at 12 MP); good daylight photos, passable indoors. Above 16 MP: rare in shipping eyewear as of 2025; likely marketing interpolation unless paired with a larger sensor module.

    Livestreaming Capable

    livestreaming_capable
    Audio & CameraUsed by 96 products

    Can the device livestream video in real time.

    The Critical View

    "Livestreaming capable" is one of the most loosely defined claims in smart glasses marketing. Some devices require a paired smartphone app to relay the stream — the glasses are merely a camera source, not an independent streaming device. Others support streaming only to one proprietary platform (e.g., Meta glasses to Instagram/Facebook) and not to arbitrary RTMP endpoints. The streaming resolution is often far below the device's recording resolution (e.g., 720p stream vs. 1080p recording). Latency, stability, and thermal throttling during sustained streams are never disclosed in spec sheets.

    How It's Measured

    Boolean. True means the device can transmit a live video feed to at least one external platform in real time, whether natively or through an official companion app. False means no supported live video output exists. Do not mark true for devices that only support screen mirroring or casting to a local display.

    Real-World Feel

    True matters primarily for content creators, field journalists, and first-person POV streamers. In practice, streaming from glasses drains the battery 2-3x faster than local recording. Most current smart glasses sustain 15-30 minutes of continuous livestreaming before thermal throttling or battery depletion. Buyers should verify which platforms are supported — a device locked to one ecosystem may not meet their needs.

    Max Recording Duration

    max_recording_duration_mins
    Audio & CameraUsed by 29 products

    Longest single video recording allowed, in minutes.

    The Critical View

    This is one of the most practically important camera specs — and one of the most hidden. Many smart glasses impose hard recording caps (often 1, 3, 5, or 10 minutes per clip) due to thermal management, battery constraints, storage limitations, or deliberate privacy design choices. Manufacturers sometimes omit this limit entirely from spec sheets, letting buyers discover it only after purchase. Some devices auto-stop recording when the processor reaches a thermal threshold, making the effective limit variable and lower than any stated maximum. The cap may also differ by resolution: a device might allow 5 minutes at 1080p but only 1 minute at 4K. Privacy-focused designs (e.g., Snap Spectacles) intentionally enforce short limits.

    How It's Measured

    Integer in minutes. Represents the longest single continuous recording the device will produce before automatically stopping, at the device's most commonly used video resolution. If the limit varies by resolution, store the value for the highest resolution that sustains at least 30 fps. A null value may mean either "no hard limit" (constrained only by storage/battery) or "not disclosed" — attempt to verify which through user reports or reviews before assuming unlimited.

    Real-World Feel

    1-3 minutes: privacy-first designs (Snap Spectacles); adequate for social clips but frustrating for events or POV documentation. 5 minutes: common in first-generation camera glasses; enough for short moments but not for recording a meeting or lecture. 10-30 minutes: the practical range for useful POV recording (walks, tours, demonstrations). 30+ minutes: rare in glasses form factor due to battery and heat; achievable by some headset-style devices. For professional use cases (field service, medical, journalism), anything under 15 minutes is a dealbreaker.

    Microphone Array

    microphone_array_type
    Audio & CameraUsed by 62 products

    Describes the microphone hardware (e.g., "dual beamforming", "3-mic array").

    The Critical View

    Manufacturers rarely standardize how they describe microphone arrays. "Dual microphone" could mean two omnidirectional mics with no beamforming (poor noise rejection) or two mics with active beamforming DSP (good noise rejection) — the spec sheet makes them sound identical. "Beamforming" itself is a software/DSP feature layered on the hardware; the same two-mic hardware can be "beamforming" or not depending on firmware. The number of microphones alone says little — a well-tuned 2-mic beamforming array often outperforms a poorly configured 4-mic array. Wind noise performance, which matters enormously for outdoor eyewear, is almost never specified.

    How It's Measured

    Stored as free text. Common values include "single omnidirectional", "dual beamforming", "3-mic array with noise cancellation", "5-mic spatial audio array". Prefer a format that includes count + type + processing: e.g., "2-mic beamforming" or "4-mic array with wind noise reduction". When manufacturers provide vague descriptions like "advanced microphone system", attempt to verify the actual mic count and configuration from teardown or FCC documentation.

    Real-World Feel

    For voice commands and phone calls in quiet environments, even a single mic suffices. For outdoor use (walking, cycling, wind), a beamforming array with 2+ mics is essential — without it, voice assistants become unusable and call quality degrades severely. For video recording, mic quality matters as much as camera quality: a high-MP camera paired with a poor single mic produces footage that sounds like it was recorded inside a washing machine. Spatial audio recording (3+ mics) is emerging for immersive content capture.

    Speaker Count

    speaker_count
    Audio & CameraUsed by 62 products

    Number of individual speaker drivers.

    The Critical View

    Most smart glasses ship with 2 speakers (one per temple arm), providing basic stereo. Some premium models include 4-6 drivers (typically a full-range driver plus a tweeter per side) for richer sound, but manufacturers sometimes inflate the count by listing passive radiators or bass ports as "speakers". A passive radiator is not an independently driven speaker — it improves bass response of an adjacent active driver but produces no sound on its own. A "4-speaker system" that is actually 2 active drivers + 2 passive radiators delivers noticeably less output than a true 4-driver system.

    How It's Measured

    Integer count of speaker drivers. Count only active, electrically driven transducers. Passive radiators, bass reflex ports, and resonance chambers should not be counted. When manufacturer specs are ambiguous (e.g., "quad speaker system"), verify through teardown data or review analysis whether all units are active drivers.

    Real-World Feel

    1 speaker (mono): adequate for notifications and voice assistant responses; no spatial separation. 2 speakers (stereo): the baseline for music and media; provides basic left-right separation given the close proximity to ears. 4+ speakers (multi-driver): diminishing returns in the glasses form factor because the drivers are millimeters from the ear. The quality of DSP tuning and driver selection matters far more than raw count in this product category.

    Speaker Type

    speaker_type
    Audio & CameraUsed by 92 products

    Speaker technology: open-ear directional, bone conduction, in-ear, etc.

    The Critical View

    Speaker type is the single most impactful audio spec for smart glasses, yet manufacturers frequently bury it or use vague terms like "premium audio" without specifying the underlying technology. The three main categories have radically different characteristics: (1) Open-ear directional speakers project sound toward the ear canal via small drivers on the temple arms — decent audio quality but audible to nearby people at moderate-to-high volumes. (2) Bone conduction transducers vibrate the skull to transmit sound — the ear canal stays completely open for situational awareness, but bass response is severely limited and audio leakage at high volume is significant. (3) In-ear/sealed speakers deliver the best audio fidelity but block ambient sound, defeating the purpose of see-through eyewear for many use cases. Some manufacturers claim "private listening" for open-ear designs, but independent tests show audio leakage is clearly audible to anyone within 1-2 meters at above 50% volume.

    How It's Measured

    Stored as free text. Normalize to standard categories: "open-ear directional", "bone conduction", "in-ear sealed", "hybrid" (combined approaches). If the manufacturer uses proprietary branding (e.g., Bose "Open Ear Audio"), note the underlying technology rather than the marketing name.

    Real-World Feel

    Open-ear directional (most common in 2024-2025 smart glasses): good for phone calls, podcasts, and navigation prompts while maintaining awareness of traffic and conversation. Not suitable for noisy commutes or private listening in quiet offices — people nearby will hear your audio. Bone conduction: best for runners and cyclists who need maximum environmental awareness; worst for music enjoyment due to weak bass. In-ear sealed: best audio quality but feels redundant with smart glasses — users seeking sealed audio typically prefer dedicated earbuds.

    Deep dive →

    Video Resolution

    video_resolution
    Audio & CameraUsed by 53 products

    Max video capture resolution and frame rate (e.g., "1080p@30fps").

    The Critical View

    Manufacturers frequently headline their highest resolution without noting the frame rate cap or stabilization penalty. A device advertised as "4K video" may only shoot 4K at 15-24 fps (choppy, unsuitable for action), while usable 30 fps is only available at 1080p. Electronic image stabilization (EIS) often forces a resolution downgrade or crop — the stabilized output may be 1080p even if the unstabilized max is 4K. Some devices also impose recording time limits at higher resolutions due to thermal throttling, making the headline spec impractical for real use.

    How It's Measured

    Stored as a text string combining resolution and frame rate, e.g., "1080p@30fps", "4K@30fps", "720p@60fps". There is no universal standard for this field; manufacturers variously report "1080p", "FHD", "1920x1080", or "2K". Normalize to the "{resolution}@{fps}fps" pattern. If multiple modes exist, store the highest practical mode (the best resolution that sustains at least 30 fps with stabilization on).

    Real-World Feel

    720p@30fps: minimum viable for video calls and casual clips; looks dated on modern screens. 1080p@30fps: the practical sweet spot for most smart glasses — good enough for social sharing and first-person POV content. 1080p@60fps: noticeably smoother motion, valuable for sports/action POV. 4K@30fps: overkill for the tiny sensors in current eyewear, but useful for cropping/reframing in post. Battery drain roughly doubles going from 1080p to 4K.

    Battery Capacity

    battery_capacity_wh
    Hardware & PowerUsed by 49 products

    Battery energy capacity measured in watt-hours.

    The Critical View

    Watt-hours is the most honest battery metric because it measures actual stored energy independent of usage profiles. However, manufacturers sometimes omit this figure entirely, listing only milliamp-hours (mAh) — which is meaningless without knowing the voltage. A "600mAh" battery could be 2.2Wh at 3.7V or 3.0Wh at 5V. Some products with external battery packs report only the glasses' internal battery, ignoring the required puck. Conversely, products like Viture Pro XR report the combined system capacity. Always verify whether the Wh figure covers the complete system needed for operation.

    How It's Measured

    Wh = (mAh x nominal voltage) / 1000. Nominal voltage for lithium-polymer cells is typically 3.7V but varies (3.8V for high-voltage cells). Capacity degrades over charge cycles — a fresh battery delivers 100% of rated Wh, but after 500 cycles expect 80%. Temperature also matters: cold weather (below 10C) can reduce effective capacity by 10-20% temporarily. FCC and UN 38.3 transport filings often contain the most accurate Wh figures since lying on regulatory documents carries legal consequences.

    Real-World Feel

    Audio-only smart glasses: 0.5-1.5Wh (tiny cells limited by temple arm volume). Tethered display glasses: 1.5-4Wh (small cells, since most power comes from the host). Standalone AR headsets: 15-25Wh (comparable to a small smartphone). For context, a typical smartphone has 15-19Wh. The physics of fitting a battery into a glasses temple arm severely limits capacity — there is roughly 3-4 cubic centimeters available, which caps practical capacity at about 1.5Wh with current lithium-polymer chemistry.

    Deep dive →

    Battery Life (Moderate Use)

    battery_life_moderate_use_hrs
    Hardware & PowerUsed by 80 products

    Battery runtime under moderate daily use, in hours.

    The Critical View

    This is the single most manipulated spec in the smart glasses industry. "Moderate use" has no standard definition. Manufacturers define it to maximize the headline number: some count "moderate" as Bluetooth audio playback at 50% volume with the display off. Others mean "standby with occasional notifications." Almost none define it as "actively using the camera, display, and AI assistant simultaneously" — which is how real users actually use the product. Tethered display glasses may claim "3 hours" but that assumes a dim display at 30% brightness. Products with companion charging cases often combine device battery + case charges into a single "up to 36 hours" number that buries the fact that each session is only 4 hours before you need to re-case.

    How It's Measured

    There is no ISO or IEC standard for smart glasses battery life testing. Contrast with smartphones where GSMArena and similar outlets have standardized tests. Each manufacturer invents its own "typical use" profile. The only reliable comparison is watt-hours (battery_capacity_wh) combined with known power draw of the display and SoC. Third-party reviews with standardized loop tests (video playback, camera recording, mixed use) are the only trustworthy source.

    Real-World Feel

    For audio-only smart glasses: 4-6 hours of music/calls is typical, 8+ hours is excellent. For tethered display glasses: 2-3 hours of active viewing is standard, 4+ hours is class-leading. For standalone AR headsets: 1-2 hours of mixed reality use is common (the Meta Quest 3 manages about 2.2 hours), 3+ hours is exceptional. For all categories: manufacturer claims typically overstate real-world endurance by 25-40%. If a spec sheet says "6 hours," expect 3.5-4.5 hours of genuine mixed use.

    Deep dive →

    Charging Case Charges

    charging_case_charges
    Hardware & PowerUsed by 49 products

    How many full charges the carrying case provides.

    The Critical View

    This number is a marketing multiplier. If a device gets 4 hours of battery life and the case provides 8 additional charges, the headline becomes "up to 36 hours total" — obscuring that each continuous session is still only 4 hours, and each recharge takes 20-75 minutes inside the case. Manufacturers always assume the case is fully charged and the glasses are completely dead (0%) for each cycle, which never happens in practice. Partial charges from a partially depleted case yield fewer total cycles. Some brands count the initial charge (glasses start at 100% out of box) as one of the case's charges. The case itself adds significant pocket weight (50-100g for audio frames, 150-300g for display glasses) that never appears in the device weight spec.

    How It's Measured

    A "full charge" means 0% to 100%. Real charging behavior is non-linear — going from 0% to 80% is faster and more efficient than 80% to 100%. Case capacity in Wh divided by device battery capacity in Wh gives theoretical maximum charges, but conversion losses (heat, voltage regulation) reduce the actual count by 10-15%. After 300-500 case charge cycles, its own battery degrades, reducing the number of device charges it can provide.

    Real-World Feel

    1-2 case charges doubles your day — fine for audio-only glasses used during a commute. 4-8 case charges means multi-day use without a wall outlet — suitable for travel. For display glasses that drain in 2-3 hours, even 4 case charges only gives you 8-12 hours of cumulative display time spread across the day with mandatory 30-60 minute recharge breaks. The case is not a power bank — you cannot use the glasses while they charge inside it (with rare exceptions like the XREAL charging case that supports pass-through).

    Charging Case Weight

    charging_case_weight_grams
    Hardware & Power

    Weight of the charging case in grams

    Charging Time

    charging_time_minutes
    Hardware & PowerUsed by 6 products

    Time to fully charge the device from empty, in minutes.

    The Critical View

    Manufacturers typically report the 0-100% time but bury the more useful metric: how long to get a usable charge. Most lithium-polymer batteries reach 80% in roughly 60% of the stated time due to the constant-current/constant-voltage charging curve — the last 20% takes disproportionately long. Some brands quote "fast charge" times that require a specific proprietary charger not included in the box. Others quote charging time from the case but not from a USB cable (or vice versa). Wireless charging, where available, is always slower than wired but may be the only option for sealed designs. The stated time assumes room temperature (20-25C) — charging in hot or cold conditions is slower and potentially damaging.

    How It's Measured

    Should specify: charge from 0-100%, the charging method (USB-C, case, wireless), and the power source (wattage of charger). No standard test conditions exist for smart glasses. Phone industry standards (USB PD, Qi) partially apply but most smart glasses use proprietary charging pins or pogo connectors. Some case-based charging systems trickle-charge at 1-2W, making "fast charge" claims impossible to verify without measuring current draw.

    Real-World Feel

    For audio-only frames: 60-90 minutes is typical for a full charge, with 15 minutes giving roughly 1-2 hours of use. For tethered display glasses: 45-90 minutes is standard. For standalone headsets: 1.5-2.5 hours is common. The practical question is: can you get enough charge during a coffee break (15 min) to last your next session? If the answer is no, the product has a meaningful usability gap in its charge cycle.

    IP Rating

    ip_rating
    Hardware & PowerUsed by 77 products

    Dust and water resistance certification (e.g., IPX4, IP54).

    The Critical View

    The "X" in ratings like IPX4 does not mean "no dust protection" — it means dust resistance was simply not tested. Manufacturers exploit this ambiguity: an IPX4-rated device may or may not survive a dusty environment, because nobody checked. More importantly, IP testing is done with fresh, room-temperature tap water on a brand-new unit — not on a 6-month-old device with micro-scratches on its seals. Sweat is saltwater and significantly more corrosive than test conditions assume. Several brands (especially Chinese OEMs) self-certify IP ratings without third-party lab validation. Always ask: was this certified by an independent lab (UL, TUV, SGS), or is it a manufacturer claim?

    How It's Measured

    IEC 60529 defines specific tests: IPX4 = splashing water from any direction for 10 minutes, IPX5 = 6.3mm water jet at 12.5 L/min for 3 minutes, IPX7 = submersion to 1 meter for 30 minutes. First digit (dust): 5 = dust-protected (limited ingress), 6 = dust-tight (no ingress). The rating is non-cumulative — IPX7 (submersion) does not guarantee IPX5 (jet spray) unless both are explicitly tested. Temperature cycling, aging, and repeated exposure are never part of the standard test.

    Real-World Feel

    IPX2 handles light drizzle only. IPX4 survives a sweaty workout or getting caught in rain — the minimum for any fitness use case. IPX5 handles a shower or heavy rain. IP54 or better is needed for dusty outdoor environments (construction, trail running). No current smart glasses product is rated for swimming (IPX8). Products with no IP rating should be treated as vulnerable to any moisture — even condensation on a humid day.

    Processor

    processor_name
    Hardware & PowerUsed by 82 products

    Primary SoC or processor chip name.

    The Critical View

    This field is one of the noisiest in the entire catalog. Manufacturers truncate, rename, or use marketing names that obscure the actual silicon. Qualcomm's AR platform branding has changed multiple times: "Snapdragon XR2 Gen 2" vs "Snapdragon AR1 Gen 1" vs "QCS6490" all refer to different chips that marketing materials sometimes conflate. Some devices list a companion app's phone processor rather than the on-device chip. Tethered display-only devices may have a simple display controller (like an HDMI bridge chip) but list no processor at all — this is accurate but confusing when comparing against standalone devices. Chinese OEMs sometimes list only the SoC family ("Qualcomm platform") without specifying the exact SKU.

    How It's Measured

    No standard format. Values range from full marketing names ("Qualcomm Snapdragon AR1 Gen 1") to bare part numbers ("QCS6490") to vague labels ("Custom SoC"). Cross-referencing with Qualcomm/MediaTek/Ambiq product pages is often required. For audio-only smart glasses, the processor may be a tiny Bluetooth SoC (e.g., Qualcomm QCC5171) that is irrelevant for computational performance comparisons.

    Real-World Feel

    For standalone AR glasses, the processor determines what apps can run and how responsive they feel. Snapdragon XR2 Gen 2 handles 6DoF tracking and spatial computing. Simpler chips like QCS4490 support camera capture and AI inference but not heavy 3D rendering. For tethered display glasses, the processor is nearly irrelevant — all rendering happens on the connected device. For audio-only frames, the Bluetooth SoC affects codec support (aptX, LC3) and call quality but not "performance" in a meaningful sense.

    RAM

    ram_gb
    Hardware & PowerUsed by 55 products

    Device RAM in GB.

    The Critical View

    RAM specs are only meaningful for standalone devices with on-board operating systems. For tethered display glasses or audio-only frames, RAM is either absent, negligible (256KB on a Bluetooth SoC), or irrelevant — yet some manufacturers list the RAM of the "recommended paired phone" in their marketing materials, which is deeply misleading. Even on standalone devices, usable RAM is always lower than the stated figure: the OS, background services, and eye-tracking daemons consume a baseline before any user app launches. A "6GB" standalone headset may have only 3-4GB available for applications.

    How It's Measured

    Should be reported as total physical RAM on the device itself, not the paired phone or compute puck. LPDDR4X vs LPDDR5 matters for bandwidth but is rarely disclosed. Manufacturers never specify how much RAM is reserved by the OS.

    Real-World Feel

    For standalone AR/MR headsets: 4GB is the bare minimum for a responsive experience, 6-8GB enables comfortable multitasking with 2-3 spatial apps, 12GB+ is needed for developer workflows or heavy 3D content. For tethered display glasses: RAM is typically 512MB-1GB for the display controller firmware — comparing this number against standalone devices is meaningless. For audio-only frames: RAM is measured in kilobytes and should not be listed in gigabytes.

    Storage

    storage_gb
    Hardware & PowerUsed by 55 products

    On-device storage in GB.

    The Critical View

    Headline storage numbers never reflect usable space. The operating system, pre-installed apps, and system partitions consume 8-15GB on standalone devices — a "128GB" headset may ship with only 110GB available. Critically, most smart glasses have no expandable storage (no microSD slot), making the initial choice permanent. Manufacturers of audio-only glasses sometimes list "32GB" when the device is really meant to store firmware and cached Bluetooth audio — users cannot access this storage for files. Some tethered display glasses have zero user-accessible storage but still list flash memory used for firmware, which inflates the spec sheet.

    How It's Measured

    Should represent total internal flash storage, not SD card slots or cloud storage bundled with a subscription. Formatted capacity is always lower than raw capacity (GiB vs GB discrepancy, ~7% less). No standard for reporting "available" vs "total" storage.

    Real-World Feel

    For standalone devices recording video: 1 minute of 1080p video uses roughly 130-200MB. A 32GB device with 20GB usable stores about 100-150 minutes of video. For standalone AR/MR headsets running apps: 64GB is tight, 128GB is comfortable, 256GB+ is needed for large game libraries. For audio-only frames: storage is irrelevant to the user experience and should not be a purchasing factor. For tethered displays: local storage is used only for firmware and settings, typically 1-4GB.

    Weight

    weight_grams
    Hardware & PowerUsed by 93 products

    Weight of the glasses/headset unit in grams.

    The Critical View

    Manufacturers routinely game this number. Tethered devices almost always exclude the cable and compute puck — a "45g" pair of AR glasses may actually require a 130g tethered controller that lives in your pocket. Some brands weigh frames without lenses or nose pads. Meta Ray-Ban lists the weight without its charging case, which is mandatory for recharging. Standalone devices sometimes omit the head strap or counterweight battery in their headline figure. Always check whether the stated weight is "glasses only" or "as worn." If a product ships with a required compute module clipped to the temple, that weight must be included — but many spec sheets quietly leave it out.

    How It's Measured

    Should be measured on a calibrated scale with all components required for basic operation attached (lenses in, nose pads on, temple tips installed). Industry has no enforced standard — some brands measure to 0.1g precision (implying lab conditions), others round to the nearest 5g. Ask: does the stated weight include prescription lens inserts if the product ships with a snap-in adapter?

    Real-World Feel

    Under 40g feels like ordinary glasses and can be worn all day. 40-50g is noticeable but tolerable for 2-4 hours of continuous use. Over 50g becomes uncomfortable after 30-60 minutes for most users, causing nose bridge pressure and ear fatigue. Over 80g causes significant nose fatigue within an hour and typically requires a counterweight or head strap. For reference, standard Ray-Ban Wayfarers weigh ~45g. Every additional gram concentrates on just two contact points: the nose bridge and ear hooks.

    Bluetooth Version

    wireless_bluetooth_version
    Platform & ConnectivityUsed by 76 products

    Bluetooth version — affects range, audio quality, multi-device support, and battery efficiency.

    The Critical View

    Manufacturers list the Bluetooth version but rarely specify which profiles are supported, which is what actually determines capability. BT 5.2 with LE Audio (LC3 codec) is a generational leap over BT 5.0 with SBC codec for audio quality — but a device can be "Bluetooth 5.2" without supporting LE Audio if the firmware doesn't implement it. Auracast (broadcast audio) requires BT 5.2+ but again is profile-dependent, not version-dependent. The version number is necessary but not sufficient. Also watch for devices that list "Bluetooth 5.3" but ship with a 5.0 companion app that bottlenecks the connection.

    How It's Measured

    Stated as the Bluetooth SIG specification version (e.g., "5.3"). Higher versions are backward-compatible. Key capability thresholds: 5.0 = baseline for modern devices, improved range over 4.2. 5.2 = LE Audio support possible, LC3 codec, multi-stream audio. 5.3 = improved power efficiency, better coexistence with Wi-Fi. 5.4 = Periodic Advertising with Responses (PAwR) for electronic shelf labels, minor consumer impact.

    Real-World Feel

    BT 5.0: expect 10m reliable range, basic audio (SBC/AAC), noticeable latency for video sync. BT 5.2+: potential for dramatically better audio quality (LC3 at 160kbps matches SBC at 328kbps), true wireless stereo without master/slave delay, and multi-device audio sharing. BT 5.3: measurably better battery life (5-10% improvement in sustained audio streaming). For smart glasses specifically, Bluetooth handles: companion app communication, audio streaming, phone notifications, and sometimes sensor data relay. Range in practice is 5-8m indoors through walls, regardless of spec sheet claims of 30m+.

    Connection Type

    connection_type
    Platform & ConnectivityUsed by 82 products

    Wired or wireless — determines latency, bandwidth, and freedom of movement.

    The Critical View

    Manufacturers often list "Wireless" when the device supports Bluetooth audio but still requires a USB-C cable for display output. XREAL Air is "wired" for display but "wireless" for nothing — yet some listings conflate its Bluetooth audio support with wireless connectivity. Conversely, some devices advertised as "wireless" use Wi-Fi Direct for display streaming with noticeable latency (30-80ms) that makes them unsuitable for gaming. The distinction between data connection and display connection is critical and almost never clarified in spec sheets.

    How It's Measured

    Binary classification: Wired or Wireless, referring to the primary display/data link. Devices that use wired for display but wireless for audio/control are classified as Wired. Devices with optional wired and wireless modes may list both. USB-C with DisplayPort Alt Mode is the dominant wired standard; Miracast/Wi-Fi Direct is the dominant wireless display standard.

    Real-World Feel

    Wired = zero perceptible latency, guaranteed bandwidth, but you are physically leashed to a device. USB-C cables degrade — expect to replace them annually with heavy use. Wireless = freedom of movement but 20-80ms added latency (fine for video, bad for gaming), potential compression artifacts, and additional battery drain on both ends. For AR overlay use cases, even 30ms of added latency causes visible misalignment with head movements. If gaming or precision work matters, insist on wired.

    Deep dive →

    Operating System

    operating_system
    Platform & ConnectivityUsed by 71 products

    The OS determines your app library, update cadence, and long-term support.

    The Critical View

    Manufacturers love coining proprietary OS names that are just Android forks with a custom launcher. "Snap OS" is a real differentiated platform; many Chinese OEM offerings labeled as proprietary are AOSP (Android Open Source Project) with locked bootloaders. The OS name alone tells you almost nothing — what matters is: (a) does it have a real app store? (b) who controls updates? (c) what is the historical update cadence? Meta Horizon OS gets regular updates; many "proprietary" systems ship and are never updated. visionOS is a walled garden with premium polish; Android XR is Google's bet on openness but is still maturing.

    How It's Measured

    Recorded as the manufacturer-stated OS. AURAI editorially notes when a "proprietary" OS is a known Android fork. Version numbers are tracked separately when disclosed. Be skeptical of "Android-based" claims that don't specify the AOSP version — some ship Android 9 in 2026.

    Real-World Feel

    visionOS = best polish, fewest apps, Apple lock-in. Android XR = growing ecosystem, fragmentation risk across OEMs. Meta Horizon OS = strong social/gaming library, Meta account required. Snap OS = excellent camera integration, tiny app ecosystem. Proprietary = expect exactly the features it ships with and nothing more. If the manufacturer goes under, a proprietary OS device becomes e-waste faster than an Android-based one.

    Deep dive →

    Software Ecosystem

    software_ecosystem
    Platform & ConnectivityUsed by 75 products

    How rich and alive is the app/software world around this device.

    The Critical View

    This is the field manufacturers inflate most aggressively. "Thousands of apps" can mean thousands of flat Android phone apps that technically launch but are unusable on a head-mounted display. Meta Quest's library of 500+ purpose-built spatial apps is categorically different from a Rokid device claiming "Android app compatibility." Developer SDK availability, active third-party development, and app store curation quality matter more than raw app count. Watch for "coming soon" ecosystems that never arrive — multiple AR glasses have launched promising app stores that remained empty.

    How It's Measured

    AURAI editorial assessment, not a standardized metric. Factors include: number of purpose-built apps (not sideloaded phone apps), SDK public availability, active developer community size, app store update frequency, and notable third-party integrations. Stored as free text describing the ecosystem state.

    Real-World Feel

    A device with 50 purpose-built spatial apps (Meta Quest, Apple Vision Pro) delivers a dramatically better experience than one with "10,000 compatible Android apps" that were never designed for the form factor. For productivity buyers: check if the specific apps you need (Zoom, Slack, Google Workspace) have native versions. For entertainment: check streaming app availability (Netflix famously withheld from Meta Quest for years). Ecosystem health is the #1 predictor of whether you'll still be using the device in 12 months.

    Deep dive →

    Supported Apps

    supported_apps
    Platform & ConnectivityUsed by 97 products

    Which specific apps work on this device — the ground truth of daily utility.

    The Critical View

    This field is where manufacturer promises meet reality most brutally. A "supported" app can mean: (a) a native spatial/glasses-optimized app, (b) a flat phone app that technically launches but is unusable with the form factor, (c) a web app accessible through the device browser, or (d) a deep integration where the glasses surface app notifications/controls without running the full app. These four tiers deliver wildly different experiences. Manufacturers frequently list apps as "supported" during product announcements and then ship without them, or apps are available at launch but removed months later (Netflix pulling from Meta Quest was a notable example). "Coming soon" apps should be treated as nonexistent until independently verified.

    How It's Measured

    JSONB array of application names. AURAI records apps only after independent verification of availability (app store listing or reliable user reports), not based solely on manufacturer press releases. The array does not distinguish between native spatial apps and sideloaded/compatibility-mode apps — editorial notes provide that context.

    Real-World Feel

    For most buyers, 3-5 specific apps determine whether a device is useful or a shelf ornament. Before purchasing, check: (1) Are YOUR must-have apps in this list? (2) Are they native or compatibility-mode? (3) When were they last updated on this platform? A device with 500 apps you don't use is less valuable than one with 5 apps you use daily. For AR display glasses: streaming apps (Netflix, YouTube, Disney+) and productivity apps (Zoom, Google Workspace) are the killer use cases. For audio glasses: music (Spotify, Apple Music) and communication (phone calls, messages) are the baseline.

    Wi-Fi Standard

    wireless_wifi_standard
    Platform & ConnectivityUsed by 73 products

    Wi-Fi generation — determines streaming bandwidth, latency, and network congestion handling.

    The Critical View

    Most lightweight smart glasses (audio, camera categories) have no Wi-Fi at all — they rely on Bluetooth and the paired phone's connection. Wi-Fi is primarily found in standalone AR/VR headsets that need high-bandwidth streaming (cloud rendering, wireless PC VR). Manufacturers sometimes list Wi-Fi capability without specifying the standard, or list the chip's maximum capability rather than what the device firmware actually enables. "Wi-Fi 6E" support is meaningless if the 6GHz band is regulatory-blocked in your country (still restricted in many markets outside the US and EU).

    How It's Measured

    Stated as Wi-Fi generation (Wi-Fi 5 = 802.11ac, Wi-Fi 6 = 802.11ax on 2.4/5GHz, Wi-Fi 6E = 802.11ax on 2.4/5/6GHz, Wi-Fi 7 = 802.11be). Key differentiators: Wi-Fi 5 = up to 3.5Gbps theoretical, adequate for 1080p streaming. Wi-Fi 6 = better in crowded environments (OFDMA), improved battery (Target Wake Time). Wi-Fi 6E = uncongested 6GHz spectrum, critical for low-latency wireless VR. Wi-Fi 7 = Multi-Link Operation, highest throughput.

    Real-World Feel

    For wireless VR/AR streaming (e.g., Meta Quest with Air Link, or cloud-rendered AR): Wi-Fi 6E on 6GHz is the minimum for a stutter-free experience — 5GHz bands in apartment buildings are often congested to the point of frame drops. For standalone devices doing web browsing and app downloads, Wi-Fi 5 is perfectly adequate. Actual throughput depends more on your router than the glasses — a Wi-Fi 7 headset on a Wi-Fi 5 router gives you Wi-Fi 5 speeds. If the device has no display (audio-only glasses), Wi-Fi is used only for firmware updates and initial setup; the standard barely matters.

    AI Features

    ai_features
    Interaction & AIUsed by 73 products

    What the device can do with AI — the features that justify calling it "smart."

    The Critical View

    AI features are the most volatile and overpromised category in smart glasses marketing. "Live Translation" can mean real-time bidirectional conversation translation (impressive, rare, requires cloud) or simple text-to-speech of pre-translated phrases (trivial). "Visual Search" can mean point-your-camera-at-anything Google Lens-style identification (Meta Ray-Ban does this well) or "scan a QR code" relabeled as AI. Nearly all AI features require cloud connectivity — if the manufacturer doesn't explicitly state "on-device," assume it needs internet and will not work in airplane mode, poor signal areas, or after the company shuts down the backend. Many AI features launch as impressive demos and degrade over time as the company reallocates server resources.

    How It's Measured

    JSONB array of feature name strings. Common values: "Live Translation", "Visual Search", "Object Recognition", "Scene Description", "Smart Summarization", "Real-time Captioning", "AI Assistant", "Contextual Reminders", "Navigation Overlay". Features are recorded based on manufacturer claims and independent verification. On-device vs. cloud-dependent is not captured in this field but is noted in editorial content.

    Real-World Feel

    The AI features that actually deliver daily value in 2026: (1) Real-time captioning/translation — transformative for hearing-impaired users and travelers, but quality varies from "usable" to "comedy of errors" depending on language pair and ambient noise. (2) Visual identification ("What am I looking at?") — Meta AI does this well with Ray-Ban Meta cameras; most others are too slow (3-5 seconds) for natural use. (3) Smart reminders and contextual awareness — still largely aspirational; no current device reliably reminds you of someone's name when you see them. If an AI feature requires you to stop, hold still, and wait more than 2 seconds for a response, you'll stop using it within a week.

    AI Models

    ai_models
    Interaction & AIUsed by 97 products

    Which specific AI brains power the device — brand names behind the intelligence.

    The Critical View

    This field emerged in 2024-2025 as manufacturers began name-dropping AI model brands for marketing cachet. "Powered by ChatGPT" can mean deep API integration with optimized prompts and device context, or it can mean "we added an OpenAI API key and wrapped it in a voice interface." The model name alone says nothing about: (a) which model version (GPT-4o vs GPT-3.5 makes an enormous difference), (b) whether it has access to device sensors (camera, location) for context-aware responses, (c) rate limits and latency, (d) what happens when the API subscription economics change and the manufacturer downgrades. Models also change without notice — a device "powered by Gemini" today might silently switch to a cheaper model tier tomorrow.

    How It's Measured

    JSONB array of AI model/platform names. Records the manufacturer-stated AI integrations. Version specifics (e.g., "GPT-4o" vs "GPT-4o-mini") are captured when disclosed but manufacturers frequently omit this detail. Multiple models can coexist on one device (e.g., on-device model for quick responses + cloud model for complex queries).

    Real-World Feel

    The AI model determines conversation quality, response speed, and capability ceiling. On-device models (small, fast, limited) handle basic commands in <500ms. Cloud models (large, powerful, latency-dependent) handle complex reasoning in 2-5 seconds. The best implementations use a hybrid approach: small model for instant acknowledgment, cloud model for actual answer, with graceful fallback when offline. For buyers: the model name matters less than the integration quality. Meta AI on Ray-Ban Meta with full camera/audio context and sub-second responses outperforms a "ChatGPT-powered" device that takes 4 seconds to respond to a voice query without any sensor context.

    Control Methods

    control_methods
    Interaction & AIUsed by 73 products

    How you interact with the device — tap, talk, wave, or point.

    The Critical View

    Manufacturers list every input method their hardware theoretically supports, but the real question is: which input method is the PRIMARY one, and how well does it actually work? A device might list "Gesture" control but only recognize two gestures reliably. "Voice" control might require cloud connectivity and fail offline. Touchpad areas on temple arms are often so small that precise interaction is frustrating. Eye tracking varies enormously in calibration reliability and works poorly for users with certain eye conditions or glasses prescriptions. The number of control methods matters less than whether the primary method is responsive enough for daily use without cursing.

    How It's Measured

    JSONB array of supported input method strings. Common values: "Voice" (microphone-based commands), "Touchpad" (capacitive touch surface on frame), "Gesture" (hand/finger tracking via cameras), "Controller" (physical handheld device), "Eye Tracking" (gaze-based selection), "Head Tracking" (look-to-select). Array indicates hardware capability, not quality of implementation.

    Real-World Feel

    Voice: works well in quiet environments, embarrassing and unreliable in public/noisy settings — plan for 30-40% failure rate on a busy street. Touchpad: the most reliable fallback but limited to swipe/tap interactions. Gesture: transformative when it works (Meta Quest hand tracking is excellent), frustrating when it doesn't (most glasses-form-factor devices lack the camera quality for reliable hand tracking). Controller: highest precision and speed but you're carrying extra hardware. Eye tracking: the fastest selection method when calibrated, but current tech needs recalibration across sessions and struggles with bright sunlight. For daily use, devices with touchpad + voice offer the best reliability-to-convenience ratio.

    Privacy Features

    privacy_features
    Interaction & AIUsed by 40 products

    What protections exist for your privacy and the privacy of people around you.

    The Critical View

    Privacy features in smart glasses are uniquely important because these devices have outward-facing cameras and always-on microphones worn in public spaces. A "Recording LED" that is tiny, dim, and invisible from more than 2 meters away is functionally useless as a social signal — but manufacturers check the "has recording indicator" box regardless. "On-device processing" is a meaningful privacy feature when it means your data never leaves the glasses, but some manufacturers use it to mean "initial processing on-device, then uploaded to cloud for refinement." End-to-end encryption claims should specify: encrypted in transit, at rest, or both? Who holds the keys? Can law enforcement compel decryption? The most important privacy feature — the ability to completely disable cameras and microphones via hardware switch — is vanishingly rare.

    How It's Measured

    JSONB array of privacy feature name strings. Common values: "Recording LED", "Hardware Camera Shutter", "Physical Microphone Disconnect", "On-Device Processing", "End-to-End Encryption", "Guest Mode", "Auto-Delete Recordings", "Privacy Zone Geofencing", "Open Source Firmware". Features are recorded based on manufacturer documentation and teardown verification where available.

    Real-World Feel

    The privacy features that actually matter for social acceptability: (1) Visible recording LED — must be bright enough to see in daylight from 3+ meters. Meta Ray-Ban Meta's LED is visible but small; Snap Spectacles' spinning LED ring was more obvious. (2) Hardware camera disable — a physical shutter or switch, not a software toggle that could be overridden. (3) On-device processing for sensitive data — biometric data, health metrics, and conversation audio should never leave the device without explicit consent. For enterprise use: data residency, GDPR compliance, and IT admin controls are non-negotiable. The sad reality: most smart glasses treat privacy features as a checkbox exercise rather than a core design principle. Buyers who care about privacy should prioritize devices with hardware-level controls over software promises.

    Deep dive →

    Sensors

    sensors
    Interaction & AIUsed by 66 products

    The hardware sensor loadout — determines what the device can perceive about you and your environment.

    The Critical View

    Sensor lists in spec sheets are typically accurate (it's hard to fake having a gyroscope), but what's misleading is the implication of capability. Having a "depth sensor" could mean a 30,000-point LiDAR array (Apple Vision Pro) or a single IR proximity sensor with 10cm range (some budget glasses). "IMU" on a spec sheet means accelerometer + gyroscope, but the quality varies by orders of magnitude — a $2 MEMS IMU drifts noticeably in minutes; a high-grade Bosch or InvenSense unit holds calibration for hours. The sensors themselves matter less than the sensor fusion software that interprets their combined output. Meta Quest's hand tracking uses standard RGB cameras but exceptional ML models; other devices with identical cameras deliver unusable tracking.

    How It's Measured

    JSONB array of sensor type strings. Common values: "Accelerometer", "Gyroscope", "Magnetometer", "Proximity", "Ambient Light Sensor", "Depth Camera", "LiDAR", "Time-of-Flight", "Barometer", "Heart Rate Monitor", "Electrodermal Activity", "EEG", "Eye Tracking Camera", "IR Camera". Array indicates presence, not sensor quality, range, or precision specifications.

    Real-World Feel

    For AR glasses: accelerometer + gyroscope is the bare minimum (enables 3-DoF). Add magnetometer for compass heading. Add outward cameras for 6-DoF SLAM and hand tracking. Add depth sensing for accurate occlusion (virtual objects correctly hidden behind real ones). For health/wellness glasses: heart rate and SpO2 sensors are increasingly common but accuracy varies wildly compared to dedicated wearables — expect +/- 5-10 BPM versus +/- 1-2 BPM on an Apple Watch. For privacy-conscious buyers: check if the sensor list includes always-on microphones or cameras, and cross-reference with the privacy_features field.

    Tracking DoF

    native_tracking_dof
    Interaction & AIUsed by 6 products

    How many axes of head/body movement the device can track — defines the AR experience tier.

    The Critical View

    This is the clearest technical dividing line in the smart glasses market, and it is frequently misrepresented. 3-DoF (three degrees of freedom) tracks head rotation only: look left/right, up/down, tilt. 6-DoF (six degrees of freedom) adds translational tracking: lean forward/back, step left/right, stand/sit. The difference is profound — 3-DoF gives you a floating screen that rotates with your head, 6-DoF gives you objects anchored in physical space that you can walk around. Some manufacturers claim "6-DoF" when they have IMU-based 3-DoF with drift-prone visual-inertial odometry that loses tracking after minutes. True 6-DoF requires outward-facing cameras with SLAM (Simultaneous Localization and Mapping), not just an accelerometer and gyroscope.

    How It's Measured

    Integer: 0, 3, or 6. 0-DoF = no spatial tracking (audio-only glasses, basic displays). 3-DoF = rotational tracking only via IMU (accelerometer + gyroscope). 6-DoF = rotational + translational tracking via inside-out camera-based SLAM or external tracking systems. Some devices offer 3-DoF natively but can achieve limited 6-DoF with an external accessory or beacon system — these are recorded as 3.

    Real-World Feel

    0-DoF: the device is a speaker, microphone, or dumb display with no spatial awareness. Fine for audio glasses and basic HUDs. 3-DoF: virtual content rotates with your head but doesn't stay anchored in space. Acceptable for watching movies (the screen follows you) but disorienting for AR overlays. Usable for seated media consumption. 6-DoF: virtual objects stay pinned to real-world locations. You can place a virtual monitor on your desk and it stays there as you move. Required for any meaningful AR experience, gaming with room-scale movement, or professional spatial computing. The jump from 3-DoF to 6-DoF is the single biggest experiential leap in the entire product category.

    Voice Assistant

    voice_assistant
    Interaction & AIUsed by 54 products

    Which voice AI you can talk to through the glasses.

    The Critical View

    This field has become the primary battleground for smart glasses differentiation since 2024. But "supports Google Assistant" can mean the glasses have a dedicated on-device integration with wake-word detection, or it can mean "if your paired phone has Google Assistant, it passes through audio." The experience difference is night and day. Meta AI on Ray-Ban Meta is deeply integrated with visual context (the camera can see what you see). Alexa on Echo Frames is audio-only but highly responsive. ChatGPT on some devices is a web API call with 2-5 second response latency. The assistant name tells you the brand; you need editorial context to know the depth of integration.

    How It's Measured

    JSONB array of voice assistant names/brands. Includes both on-device integrations and passthrough/relay integrations without distinction in the data field. AURAI editorial notes indicate integration depth: "native" (dedicated on-device pipeline), "companion relay" (passes to phone), or "API" (cloud call with higher latency).

    Real-World Feel

    Native integrations (Meta AI on Ray-Ban Meta, Siri on Apple Vision Pro) respond in under 1 second and can access device-specific features (camera, display, sensors). Companion relay integrations add 0.5-2 seconds of latency and cannot access glasses-specific hardware. API-based integrations (ChatGPT on various devices) add 2-5 seconds and require active internet. For real utility, what matters is: can the assistant (a) respond quickly enough for conversational use? (b) access the glasses' camera for visual questions? (c) work offline for basic commands? As of 2026, only Meta AI and Apple's Siri deliver all three on their respective platforms.

    Deep dive →

    Manufacturer URL

    manufacturer_url
    Commerce & ContextUsed by 95 products

    Direct link to the manufacturer's official product page.

    The Critical View

    Manufacturer URLs are surprisingly volatile. Companies rebrand, restructure their websites, or quietly remove product pages when devices are discontinued or replaced by newer models. A dead manufacturer_url is often the first signal that a product has been silently end-of-lifed. Some manufacturers also maintain different URLs for different regions, and the "official" page may not reflect pricing or availability in the user's market. AURAI monitors these links for freshness and flags 404s as a potential discontinuation signal.

    How It's Measured

    We store the canonical English-language product page URL. For brands with region-specific domains (e.g., .cn vs .com), we prefer the international or US-facing page. We do not store campaign-tracked URLs (no UTM parameters). If a manufacturer has separate pages for each SKU/color variant, we store the parent product page that lists all variants.

    Real-World Feel

    This link is your starting point for official specs and purchasing information, but be aware that manufacturer product pages are marketing documents, not technical specifications sheets. Cross-reference critical specs against FCC filings, teardown reports, and independent reviews.

    Marketing Tags

    marketing_tags
    Commerce & ContextUsed by 97 products

    Manufacturer marketing labels and product highlight keywords.

    The Critical View

    Marketing tags are captured directly from manufacturer materials and are inherently promotional. We include them because they reveal how the manufacturer positions the product and what features they consider differentiating. However, treat these as claims, not verified facts. "Military-grade durability" might mean it survived a single drop test in a lab. "All-day battery" might mean 4 hours of the lightest possible usage. "Prescription-ready" might mean a $200 aftermarket lens adapter exists, not that you can order prescription lenses from the manufacturer. AURAI stores these tags for searchability and context but does not editorially endorse them. Our verified specs in other fields are the source of truth.

    How It's Measured

    Captured as-is from official marketing materials, press releases, and product packaging. Stored as a JSONB array of strings. We do not editorially curate or verify marketing tags — they are preserved as primary-source marketing claims. Duplicate or near-duplicate tags across products are normalized for consistent filtering (e.g., "Bone Conduction" not "bone-conduction speakers").

    Real-World Feel

    Marketing tags are useful for discovering products that claim specific features, but always cross-reference against the verified spec fields. If a product is tagged "Spatial Audio" but the speaker_type field says "mono directional," the spec field is authoritative. Use marketing_tags for broad discovery and the structured spec fields for purchasing decisions.

    MSRP (USD)

    msrp_usd
    Commerce & ContextUsed by 91 products

    Official launch price in USD for cross-catalog comparison.

    The Critical View

    MSRP is the single most gamed number in consumer electronics. Common tactics in smart glasses: (1) Announcing a low MSRP for the base model that ships without a display or with minimal storage, then charging 40-60% more for the configuration most people actually want. (2) Quoting MSRP before tax in markets where tax is included in the sticker price, making US prices look artificially lower than EU equivalents. (3) Setting MSRP high at launch then running permanent "sales" to create an illusion of value. (4) Excluding required accessories from the price: AR glasses that cost $399 but require a $149 compute puck are really $548 devices. (5) Regional pricing disparities where the same product costs $299 in the US but EUR 399 in Europe, far exceeding currency conversion. AURAI stores the USD MSRP as a consistent benchmark but our editorial notes call out hidden costs, required accessories, and the real street price when it diverges significantly.

    How It's Measured

    We record the manufacturer's official USD MSRP at the time of general availability launch. We do not update this field when prices drop later (that is tracked in regional_info). For products not officially sold in USD, we use the manufacturer's own USD conversion if provided, or the exchange rate at launch date. Tax is never included. Required accessories are noted separately, not added to MSRP.

    Real-World Feel

    The MSRP tells you what the manufacturer thinks the product is worth, not what you will actually pay. Check regional_info for current street prices. A $299 MSRP on audio glasses versus a $2,199 MSRP on a spatial computing headset does not mean one is "cheaper" — they serve entirely different purposes. Compare MSRP within the same product category for meaningful price evaluation.

    Release Date

    release_date
    Commerce & ContextUsed by 69 products

    Official release or first-available date.

    The Critical View

    Release dates in the smart glasses industry are notoriously unreliable. Manufacturers routinely announce aspirational dates at CES or MWC, then quietly push them back. Some brands count the date of a limited developer preview as the "release date" even though general consumers cannot buy the product for months afterward. Others use regional staggering to claim an early release date based on a single-market launch (often China) while the product remains unavailable in the US or EU for 3-6 months. AURAI records the date a product became genuinely purchasable by a regular consumer, not the date of a keynote announcement or an invite-only early access program.

    How It's Measured

    We use the ISO 8601 date format (YYYY-MM-DD). When only a month or quarter is known, we store the first day of that period and note the precision in editorial copy. Pre-order open dates are not treated as release dates. For phased regional launches, we record the earliest date any consumer market had open retail availability.

    Real-World Feel

    A product with a release date six months in the future is speculation, not a purchase decision. Even products with confirmed dates slip. If you are deciding between a device available today and one "releasing next quarter," weight the bird in hand heavily.

    Social Stealth Score

    social_stealth_score
    Commerce & ContextUsed by 64 products

    Editorial 1-10 rating of how discreet the device looks in public.

    The Critical View

    This is AURAI's proprietary editorial score, not a manufacturer claim. We created this metric because the binary looks_like_regular_glasses field does not capture the spectrum between "invisible" and "cyborg cosplay." A score of 8-10 means you could wear it to a job interview. A 4-6 means you will get curious looks but not hostile reactions. A 1-3 means you look like you are beta-testing for a tech company, which may be fine for your use case but will get you asked to remove the device in many restaurants and offices. Manufacturers have no incentive to honestly rate social acceptability, so this score is entirely independent editorial judgment.

    How It's Measured

    Scale of 1 (maximum attention-drawing, e.g., bulky VR headset) to 10 (indistinguishable from fashion eyewear). Factors weighted: temple thickness vs. normal glasses (30%), visible cameras/sensors (25%), front-face profile normalcy (20%), weight-induced fit issues visible to others (15%), and LED/indicator visibility (10%). Scored by editorial team using real product images and hands-on assessment. Updated when new hardware revisions change the industrial design.

    Real-World Feel

    A social_stealth_score of 7+ means you can wear it daily without it becoming a conversation topic. Below 5, expect the device to define your social interactions ("What are those things on your face?"). For enterprise use in customer-facing roles, we recommend 8+. For personal daily wear, 6+ is the practical minimum. For home/office-only use, the score is irrelevant — focus on specs instead.

    Deep dive →

    Status

    status
    Commerce & ContextUsed by 97 products

    Product lifecycle stage (Rumored, Announced, Available, Discontinued).

    The Critical View

    Status is one of the most manipulated fields in consumer electronics. Manufacturers announce products years before shipping to freeze competitors out of the market ("vaporware"). Some devices sit in "Announced" limbo for 18+ months, never reaching production. Others are technically "Available" but only through a developer program with a $1,500 waitlist. AURAI tracks the gap between announcement and actual retail availability because some brands use perpetual "coming soon" status as free marketing. Watch for products marked "Available" that are only purchasable in one country or through a single obscure retail channel.

    How It's Measured

    We track four discrete states. "Rumored" means credible leaks or regulatory filings exist but the manufacturer has not confirmed. "Announced" means an official reveal has occurred. "Available" means a consumer can place an order and receive the product today, not a pre-order or developer kit. "Discontinued" means the manufacturer has ceased production. We do not use "Prototype" as a public-facing status because nearly every company calls their unreleased product a prototype.

    Real-World Feel

    If a product shows "Announced," do not plan your workflow around it. History shows roughly 30% of announced smart glasses never ship to consumers or ship with significantly reduced specs. Only "Available" means you can actually buy and use the device today.

    Deep dive →

    Target Personas

    target_personas
    Commerce & ContextUsed by 97 products

    Intended user profiles (e.g., "The Executive," "The Creator," "The Athlete").

    The Critical View

    Target personas are derived from a combination of manufacturer marketing and AURAI editorial analysis. Manufacturers tend to claim their product is "for everyone," which is never true. A $2,000 spatial computing headset is not for casual consumers regardless of how the ad shows a grandmother using it. AURAI assigns personas based on the realistic intersection of the device's capabilities, price point, form factor, and ecosystem requirements. We distinguish between who the manufacturer wants to sell to (aspirational marketing) and who will actually benefit from the product (practical reality). A device with no display but excellent audio is tagged for "The Podcast Listener" and "The Hands-Free Professional," not "The AR Developer."

    How It's Measured

    Stored as a JSONB array of string labels from a controlled vocabulary. Personas are assigned editorially, not algorithmically, though we use spec profiles to suggest initial assignments. Each product typically receives 2-4 persona tags. We avoid overly broad personas ("tech enthusiast") in favor of specific use-case-driven profiles.

    Real-World Feel

    Use persona tags to quickly filter for devices that match your actual use case. If you are a cyclist who wants turn-by-turn navigation, filtering for "The Athlete" or "The Commuter" will surface relevant devices faster than reading through spec sheets. But remember that persona tags reflect our editorial judgment of the best fit — your specific needs may cross persona boundaries.

    Use Cases

    use_cases
    Commerce & ContextUsed by 97 products

    Primary usage scenarios (e.g., "Teleprompting," "Navigation," "Translation").

    The Critical View

    Use cases sit at the intersection of manufacturer intent and real-world capability. A product may be marketed for "Real-time Translation" but the translation feature requires a constant smartphone connection, a paid subscription, and only supports 5 languages — that is a very different "translation" device than one with onboard processing and 40 languages. AURAI assigns use cases based on what the product can credibly do today, not what the manufacturer roadmap promises. We are particularly skeptical of use cases that require software features not yet released or that depend on third-party app ecosystem development. If a use case requires asterisks and footnotes, we either do not list it or we add a qualifying note.

    How It's Measured

    Stored as a JSONB array of strings from a semi-controlled vocabulary. We distinguish between primary use cases (the product genuinely excels at this) and secondary use cases (it can technically do this but other products do it better). Only primary use cases are stored in this field. Editorial team validates use cases against hands-on testing, independent reviews, and user community feedback. Use cases are revisited when major firmware updates add or remove functionality.

    Real-World Feel

    If your purchase decision hinges on a specific use case ("I need this for cycling navigation"), check whether that use case appears in this array. If it does not, the product likely was not designed with that scenario in mind, even if it could theoretically work. Products with fewer, more focused use cases often execute better than "Swiss Army knife" devices that claim to do everything.

    Announced Date

    announced_date
    Identity & ClassificationUsed by 6 products

    Date product was first publicly announced (distinct from release_date)

    Brand ID

    brand_id
    Identity & ClassificationUsed by 97 products

    Links this product to its manufacturer in the brands table.

    Certifications

    certifications
    Identity & ClassificationUsed by 97 products

    Regulatory certifications array, e.

    Computational Locus

    computational_locus
    Identity & ClassificationUsed by 88 products

    Where the processing brain lives — on the glasses or somewhere else.

    The Critical View

    This is the single most important architectural decision a manufacturer makes, yet it is routinely buried or obscured. "Standalone" and "all-in-one" are used interchangeably by marketing teams even when the device offloads heavy compute to a phone. Qualcomm AR1/AR2-based glasses that require a paired smartphone are sometimes marketed as "standalone" because they have an onboard chip — but that chip only handles sensor fusion, not application logic. Apple Vision Pro is genuinely onboard; XREAL Air 2 is genuinely tethered. Meta Ray-Ban Stories occupy a grey zone: onboard for audio, paired smartphone for AI features. Always check what happens when the companion device is absent.

    How It's Measured

    No ISO standard. AURAI classifies into four tiers: (1) Onboard/Integrated — runs apps independently (e.g., Apple Vision Pro, Meta Quest 3). (2) Dedicated Compute Unit — separate puck or belt pack (e.g., Magic Leap 2 Compute Pack). (3) Paired Smartphone — requires phone for app logic, glasses handle display/sensors (e.g., XREAL Air, RayNeo Air 2). (4) Tethered Host Device — wired to PC/console (e.g., Lenovo Legion Glasses). Some devices span multiple tiers depending on mode.

    Real-World Feel

    Onboard means true portability but shorter battery life (2-3 hrs typical) and more heat/weight on your face. Tethered gives better performance and battery endurance but kills mobility. Paired Smartphone is the current sweet spot for lightweight glasses but ties you to phone battery drain and Bluetooth latency. If you want to use glasses on a plane without pulling out a laptop, only Onboard devices qualify.

    Deep dive →

    Contrast Ratio

    contrast_ratio
    Identity & ClassificationUsed by 8 products

    Display contrast ratio, e.

    Fov Horizontal

    fov_horizontal
    Identity & ClassificationUsed by 2 products

    Horizontal field of view in degrees

    Fov Vertical

    fov_vertical
    Identity & ClassificationUsed by 2 products

    Vertical field of view in degrees

    Generation

    generation
    Identity & ClassificationUsed by 74 products

    Numeric generation/version identifier (e.g., 1, 2, 3).

    The Critical View

    Generation numbering is straightforward in theory but manufacturers do not follow consistent rules. Some brands increment by 1 for each annual refresh (Gen 1, Gen 2, Gen 3). Others skip numbers for marketing impact (jumping from Gen 2 to Gen 5 to imply a larger leap). Some use decimal versions (1.5, 2.1) for mid-cycle refreshes, though AURAI stores only integer generations. The generation number appended to the product name (e.g., "ProductName (Gen 2)") helps distinguish products in the same family, but it is not a reliable indicator of how much the product has actually improved — a "Gen 3" might be a minor spec bump while a "Gen 2" from a competitor might be a ground-up redesign.

    How It's Measured

    Stored as a positive smallint. Null means the product is either the first (and only) generation or the manufacturer does not use generation numbering. The generation value feeds into the computed name column: when set, the display name becomes "Brand Model (Gen N)." We do not store sub-versions or decimal revisions.

    Real-World Feel

    A higher generation number within the same product line generally means newer hardware, but do not assume Gen 3 is three times better than Gen 1. Check the actual spec differences. Sometimes a brand's Gen 1 product from last year outperforms a competitor's Gen 3 if the competitor has been making incremental updates to a weaker base design.

    Has Audio Jack

    has_audio_jack
    Identity & ClassificationUsed by 97 products

    Whether device has a 3.

    Has Haptics

    has_haptics
    Identity & ClassificationUsed by 97 products

    Whether device has haptic feedback

    ID

    id
    Identity & ClassificationUsed by 97 products

    Internal unique identifier for the product.

    Ipd Max mm

    ipd_max_mm
    Identity & ClassificationUsed by 8 products

    Maximum interpupillary distance in mm

    Ipd Min mm

    ipd_min_mm
    Identity & ClassificationUsed by 8 products

    Minimum interpupillary distance in mm

    Lens Width mm

    lens_width_mm
    Identity & ClassificationUsed by 1 product

    Lens width in millimeters (lens-specific dimension, distinct from overall product dimension_width_mm)

    Looks Like Regular Glasses

    looks_like_regular_glasses
    Identity & ClassificationUsed by 97 products

    Does the device pass as normal eyewear to casual observers?

    The Critical View

    This is a subjective editorial judgment, and we own that openly. Manufacturers universally claim their products "look like regular glasses," but the bar varies wildly. Meta Ray-Ban Stories genuinely pass in most social situations. A device with a visible camera bump, LED ring, or temples twice the width of normal glasses does not pass, even if the marketing renders digitally slim down the temple thickness by 30%. AURAI evaluates this based on real-world product photography (not rendered marketing images) and physical hands-on assessment when available. We apply a "would a stranger on the subway notice?" test, not a "could this theoretically be mistaken for glasses in a dark room?" test.

    How It's Measured

    Boolean: true or false. True means the device passes casual inspection at normal social distance (1-2 meters) in a well-lit environment. We consider: temple thickness relative to fashion eyewear norms, presence of visible cameras/sensors/LEDs, overall weight distribution visible from the front, and whether the device requires any visible accessories (cables, belt packs). Promotional renders are not used for this assessment — only real product photos or hands-on evaluation.

    Real-World Feel

    If this field is true, you can wear the device in social and professional settings without drawing attention. If false, expect questions, stares, or outright bans in privacy-sensitive environments. For many buyers, this boolean is the single most important field in the entire database because it determines whether the device becomes a daily driver or a drawer ornament.

    Manual URL

    manual_url
    Identity & ClassificationUsed by 16 products

    URL to the official user manual (PDF or Web)

    Model Name

    model_name
    Identity & ClassificationUsed by 97 products

    Manufacturer model name, without brand prefix.

    The Critical View

    Model naming in the smart glasses industry is a chaotic mess. Common manufacturer tactics: (1) Reusing model names across generations with only a suffix change, making it easy to confuse the 2024 and 2026 versions (e.g., "Air" vs "Air 2" vs "Air 2 Pro" vs "Air 2 Ultra"). (2) Using regional model name variants where the same hardware ships as "Model X" in China and "Model Y" internationally. (3) Quiet mid-cycle hardware revisions under the same model name, where the "Glasses V2" you buy in March has different internals than the one reviewed in January. AURAI stores the internationally-marketed English model name and tracks known regional aliases in brand_naming_conventions. The model_name combined with brand name forms the computed products.name column.

    How It's Measured

    Free text, stored without the brand prefix. The brand name is joined from the brands table to compute the full product name. Model names are stored as the manufacturer officially writes them, preserving capitalization and spacing (e.g., "Air 2 Pro" not "air-2-pro"). The slug field provides the URL-safe normalized version.

    Real-World Feel

    When searching for reviews, accessories, or replacement parts, use the exact model_name as stored here. Searching for "XREAL Air" will return different results than "XREAL Air 2" — a distinction that matters when the products have fundamentally different capabilities.

    Model Number

    model_number
    Identity & ClassificationUsed by 83 products

    Internal/regulatory identifier (e.g., FCC ID, part number).

    The Critical View

    Model numbers are the most reliable way to identify exactly which hardware revision you are dealing with, because manufacturers cannot easily change them without new regulatory filings. When a model_name like "Smart Glasses Pro" could refer to three different hardware revisions sold over two years, the model_number is unambiguous. AURAI uses model numbers to detect silent hardware revisions and to cross-reference FCC/CE filings for independent spec verification. However, some manufacturers use different model numbers for different regional SKUs of identical hardware, and some low-volume brands skip formal regulatory filing entirely, leaving this field null.

    How It's Measured

    Sourced primarily from FCC filings, CE certification databases, and device packaging. Stored as-is including any alphanumeric formatting. Unique per brand (enforced by database constraint). Not all products have a discoverable model number — this field is nullable for products where no regulatory filing or packaging reference has been identified.

    Real-World Feel

    Model numbers are essential for warranty claims, finding compatible accessories, and verifying that the product you received matches what was reviewed. If you are buying secondhand, ask for the model number to confirm you are getting the exact revision you expect.

    Name

    name
    Identity & ClassificationUsed by 97 products

    The full, combined name of the product (Brand + Model Name).

    Ppd

    ppd
    Identity & ClassificationUsed by 2 products

    Pixels per degree (average)

    Price Segment

    price_segment
    Identity & Classification

    General categorization of the product's target market based on price (Budget, Mainstream, Premium, Ultra-premium).

    Slug

    slug
    Identity & ClassificationUsed by 97 products

    URL-friendly version of the full product name, used for routing and SEO (e.

    Tracking Type

    tracking_type
    Identity & ClassificationUsed by 6 products

    Tracking method description, e.

    Bridge Width

    frame_bridge_mm
    Physical DimensionsUsed by 63 products

    Nose bridge width in millimeters.

    The Critical View

    Bridge width determines whether the glasses sit correctly on your nose or slide down constantly. This measurement is critical but routinely omitted from smart glasses spec sheets — many brands list none of the traditional eyewear measurements, making fit assessment impossible before purchase. Smart glasses bridges must accommodate not just the nose but sometimes also proximity sensors, LED indicators, or IR emitters mounted on the bridge pad area. An 18mm bridge on smart glasses may feel different from 18mm on traditional eyewear because the weight distribution is different (heavier temples pulling backward). Some products use adjustable nose pads (silicone, spring-loaded) while others have a fixed molded bridge — the latter is far less forgiving of individual nose anatomy.

    How It's Measured

    Standard optical measurement: the horizontal distance between the inner edges of the two lenses at the narrowest point of the bridge. Measured in millimeters. In traditional eyewear, this is the middle number in the "52-18-140" notation. Smart glasses manufacturers do not always follow this convention — some measure the full bridge span including the pad arms, inflating the number by 2-4mm.

    Real-World Feel

    16-18mm: fits narrower nose bridges, common in Asian-fit and smaller frames. 19-21mm: the most common range, fits the majority of adult noses. 22-24mm: wide bridge, for larger faces or users who find standard frames pinching. If your current glasses have a bridge measurement, match it within 1mm for smart glasses. A bridge too narrow causes pinching and red marks; too wide causes the glasses to slide down, which is especially problematic for display glasses where even 2mm of vertical shift can move the display out of your visual sweet spot.

    Depth

    dimension_depth_mm
    Physical DimensionsUsed by 66 products

    Maximum device depth in millimeters (front to back).

    The Critical View

    Depth is the dimension that most betrays a device as "smart glasses" rather than regular eyewear. Normal glasses lenses protrude 8-12mm from the face. Smart glasses with embedded displays, cameras, or waveguides can extend 15-25mm, creating a visible "snout" that is the primary aesthetic giveaway. Manufacturers almost never prominently list depth because a large number hurts the "looks like regular glasses" narrative. When they do list it, they may measure from the frame front to the back of the temple hinge rather than from the frontmost lens surface to the back of the temple tip, excluding the lens protrusion. For folding glasses, the "depth" when folded (for case fit) is a different and also useful measurement that is rarely provided.

    How It's Measured

    Front-to-back distance at the deepest point. For glasses: from the front surface of the lens to the rearmost point of the frame (typically the hinge or where the temple meets the face). For headsets: from the front of the visor to the rear of the head mount. No standard reference plane — some measure perpendicular to the face, others along the temple axis. This inconsistency can create 5-10mm discrepancies between brands measuring the same physical size.

    Real-World Feel

    Under 15mm depth is visually indistinguishable from regular eyewear at conversational distance. 15-20mm is noticeable up close but acceptable in most social settings. Over 20mm is conspicuously a tech device and will draw attention. For AR display glasses, depth is often dictated by the waveguide or birdbath optical stack — thinner optics are a major engineering differentiator. For camera-equipped glasses, the camera module is often the deepest protrusion point. Depth also affects case size: bulkier glasses need bigger cases, which affects everyday portability.

    Frame Shape

    frame_shape
    Physical DimensionsUsed by 73 products

    Frame style category (Wayfarer, Aviator, Shield, Round, etc.).

    The Critical View

    Frame shape drives social acceptability more than any other spec. A technically superior device in a bulky shield form factor will be left in a drawer if it looks conspicuously "techy." Manufacturers increasingly use established eyewear terminology (Wayfarer, Aviator, Cat-Eye) to signal normalcy, but smart glasses frames are almost always thicker than their traditional counterparts due to battery and electronics in the temples. "Wayfarer-style" on a smart glasses spec sheet means "inspired by Wayfarer proportions" — not that it will pass for a regular pair of Ray-Bans at conversational distance. Shield-style frames (single continuous lens) are more common on display-equipped glasses because they accommodate larger optical assemblies, but they signal "tech device" immediately.

    How It's Measured

    No standard taxonomy. "Wayfarer" and "Aviator" are Ray-Ban trademarks used generically. Industry uses terms inconsistently: "Sport" vs "Shield" vs "Wrap" often describe the same shape. We normalize to: Wayfarer, Aviator, Round, Cat-Eye, Shield, Sport Wrap, Rectangle, Clubmaster, and Custom/Unique.

    Real-World Feel

    Wayfarer and Round frames blend into everyday life best — colleagues may not notice the tech. Aviator frames work well for outdoor use but the metal bridge can interfere with nose-mounted sensors. Shield/Sport Wrap frames are most functional for AR displays (maximum optical area) but most conspicuous in social settings. For prescription lens compatibility, Rectangle and Wayfarer shapes have the widest range of Rx insert options.

    Frame Size

    frame_size
    Physical DimensionsUsed by 72 products

    Frame size classification (Small, Standard, Large).

    The Critical View

    Frame sizing in smart glasses is far less standardized than traditional eyewear. In optical retail, frame sizes map to precise measurements (lens width + bridge + temple). In smart glasses, "Standard" is whatever the manufacturer decided to make — and most make exactly one size, labeling it "Standard" to avoid alienating potential buyers. This is a major accessibility problem: users with narrow faces find "Standard" frames sliding down their nose, while users with wide faces experience temple pressure headaches. Few smart glasses brands offer multiple size options (Meta Ray-Ban is a notable exception with Standard and Large). "One size fits all" effectively means "one size fits medium adult male faces."

    How It's Measured

    Traditional eyewear sizing uses the box system: lens width (46-58mm), bridge width (14-24mm), temple length (135-155mm). Smart glasses manufacturers may or may not follow this convention. When they do provide measurements, they should be cross-referenced with frame_bridge_mm and frame_temple_mm for actual fit assessment. The size label alone (S/M/L) is insufficient for purchase decisions.

    Real-World Feel

    Small: fits face widths under 130mm, typically women's and youth sizing. Standard: fits face widths 130-145mm, the most common adult range. Large: fits face widths over 145mm. If only one size is offered and your face width is outside 135-142mm, expect fit issues. Poor fit is the #1 reason smart glasses are returned — more than battery life, more than features. Always check the bridge and temple measurements, not just the size label.

    Height

    dimension_height_mm
    Physical DimensionsUsed by 69 products

    Maximum device height in millimeters.

    The Critical View

    Height is less critical for fit than width but matters for aesthetics and field of view. For glasses-style devices, height is the vertical measurement of the lens/frame front — taller lenses look more conspicuous but may accommodate larger displays or waveguides. Manufacturers of AR display glasses sometimes report only the lens height (relevant for optics) rather than the full frame height including the brow bar and lower rim, making devices appear sleeker on paper. For headset-style devices, height includes the full visor assembly and forehead padding, which is not comparable to glasses-style measurements.

    How It's Measured

    Vertical distance from the highest point to the lowest point of the device as worn. For glasses: typically the frame front height at the tallest lens point. For headsets: from the top of the headband mount to the bottom of the cheek rest or light blocker. Consistency across manufacturers is poor — some measure the lens opening, others the full frame including decorative elements.

    Real-World Feel

    For glasses-style devices: 35-42mm height looks proportional to most faces and resembles normal eyewear. Over 45mm starts to look like safety goggles or a visor. For display glasses specifically, a taller lens area can mean a larger eye box (the zone where the display is visible), which improves usability — you don't need to position the glasses as precisely on your nose. For headsets: height is less meaningful as a standalone number; what matters is how the weight distribution interacts with the forehead and cheek contact points.

    Temple Length

    frame_temple_mm
    Physical DimensionsUsed by 64 products

    Temple arm length in millimeters.

    The Critical View

    Temple length is where smart glasses diverge most dramatically from traditional eyewear. Normal glasses temples are 135-145mm and weigh 5-8g each. Smart glasses temples are 145-160mm and weigh 15-30g each because they contain batteries, speakers, microphones, touchpad sensors, and sometimes cameras. The extra length is needed to house these components, but it changes the leverage dynamics: a longer, heavier temple applies more downward force on the ear and more forward torque on the nose bridge. Some manufacturers use the same temple length as traditional eyewear but make the temple arms noticeably thicker (10-14mm vs the normal 4-6mm), which can interfere with hats, helmets, headbands, and over-ear headphones.

    How It's Measured

    Standard optical measurement: from the center of the hinge barrel to the end of the temple tip, following the curve. In traditional eyewear, this is the third number in the "52-18-140" notation. Smart glasses manufacturers should measure the same way but sometimes report the straight-line length rather than the along-the-curve length, understating the actual arm by 5-10mm on curved designs.

    Real-World Feel

    135-140mm: shorter arms, better for smaller heads but may not hook behind the ear securely on larger heads — risky for active use. 140-150mm: standard range, fits most adults. 150-160mm: longer arms common on tech-heavy smart glasses, may extend too far past the ear on smaller heads, causing the tips to flare outward. For smart glasses specifically: temple thickness matters as much as length. Ask whether the temples will fit under a bike helmet or alongside earbuds.

    Width

    dimension_width_mm
    Physical DimensionsUsed by 68 products

    Maximum device width in millimeters (temple tip to temple tip).

    The Critical View

    For glasses-form-factor devices, width is typically measured from the outside of the left temple hinge to the outside of the right temple hinge (the "frame front width"). This directly determines face fit — more reliably than the "frame_size" label. However, some manufacturers measure width at the widest point of the temple arms when they bow outward, which can overstate the effective width by 5-15mm. For headset-form-factor devices (like ski-goggle-style AR headsets), width includes the full housing and may not be comparable to glasses-style measurements at all. Some spec sheets list the folded width (temples collapsed) rather than the worn width, which is useless for fit assessment.

    How It's Measured

    Should be the maximum horizontal extent of the device as worn, temple-to-temple. For glasses: measured across the frame front at the hinge points. For headsets: measured across the widest point of the chassis. Units in millimeters. There is no universal agreement on whether to measure at the hinges, at the temple bows, or at the widest point of any protruding component (camera bump, hinge mechanism).

    Real-World Feel

    For glasses-form-factor: 130-138mm suits narrow faces, 138-148mm fits average faces, 148-155mm accommodates wider faces. If the frame width is more than 5mm narrower than your face width, expect temple pressure headaches. If more than 5mm wider, the glasses will feel loose and shift during head movement — catastrophic for display glasses that require precise optical alignment. For reference, most traditional adult eyewear frames are 135-145mm wide.