Free tool
Magic Leap 1 vs Magic Leap 2
Both devices below are described from their manufacturers’ own specification pages. Where a row does not apply to one of them, it says so rather than showing a dash — that is a different statement from “the maker does not publish it”.
Magic Leap · Enterprise
Magic Leap · Enterprise
Differences you can see
Pictures first, table second. Every chart is computed from the same published fields you can read below it — nothing here is an illustration.
- Magic Leap 1
- Magic Leap 2
Resolution per eye
Panels at true pixel scale, one inside the other.
The sharpest panel here has 106% more pixels per eye than the softest — that is what you gain on small text inside a scene.
Field of view
Seen from above. The grey area is roughly the human binocular field (~200°).
- Magic Leap 270° diagonal
- Magic Leap 150° diagonal
The widest field here is 70°, the narrowest 50°. A pair of human eyes covers about 200° — even the best of these shows less than two thirds of that.
Weight
On the whole catalogue’s scale, from 30 g to 830 g.
Magic Leap 1316 g*
Magic Leap 2260 g
30 g830 g
Battery
On the whole catalogue’s scale, from 45 min to 540 min; a handful of outliers, up to 2880 min, sit beyond it.
Magic Leap 1180 min
Magic Leap 2210 min
45 min540 min
On sale, and supported
Two different questions: how long you can still buy it new, and how long the manufacturer keeps updating it. The dashed line is the end of a three-year training programme (2029).
Magic Leap 1 sold 2018–2024 · support until 2024
Magic Leap 2 sold 2022–2026 · support until 2027
available newmanufacturer support after the last sale
20172030
Sold and supported are not the same thing. A model can leave the shelves while its maker keeps shipping software to the units you already own — and it can stay on the shelves with no published support commitment at all.
What actually differs
Computed from the two records, most decisive first.
Third-party apps
- Magic Leap 1
- Vendor SDK only
- Magic Leap 2
- Ordinary Android app (APK)
Launch price
- Magic Leap 1
- $2295
- Magic Leap 2
- —
Resolution per eye
- Magic Leap 1
- 1280x960
- Magic Leap 2
- 1440x1760
Field of view
- Magic Leap 1
- 50° diagonal
- Magic Leap 2
- 70° diagonal
Brightness
- Magic Leap 1
- —
- Magic Leap 2
- 2000 nits
Refresh rate
- Magic Leap 1
- 122 Hz
- Magic Leap 2
- 120 Hz
Weight
- Magic Leap 1
- 316 g
- Magic Leap 2
- 260 g
Battery
- Magic Leap 1
- 180 min
- Magic Leap 2
- 210 min
Kiosk mode / MDM
- Magic Leap 1
- No
- Magic Leap 2
- Yes
Availability
- Magic Leap 1
- Discontinued in 2024
- Magic Leap 2
- Discontinued in 2026
Announced
- Magic Leap 1
- 2017-12-20
- Magic Leap 2
- 2021-02-01
Released
- Magic Leap 1
- 2018
- Magic Leap 2
- 2022
They match on 21 further specifications.
Full specification, row by row
39 of 84 rows differ.
| Specification | Magic Leap 1 Magic Leap | Magic Leap 2 Magic Leap |
|---|---|---|
| ManufacturerWho makes it — and who you buy support from. | Magic Leap | Magic Leap |
| CategoryWhich kind of product this is. It decides which rows below make sense at all. | AR — spatial | AR — spatial |
| Device typeWhere the computing happens, and therefore what you need to run it. | AR with a compute puck | AR with a compute puck |
| AnnouncedThe day it was shown, which is not the day it shipped. | 2017-12-20 | 2021-02-01 |
| ReleasedYear the model went on sale. Older is not automatically worse in this category. | 2018 | 2022 |
| AvailabilityWhether you can still order it new today. | Discontinued in 2024 | Discontinued in 2026 |
| Support endsThe date the manufacturer stops supporting it. Decisive for a fleet you are about to buy. | 2024-12-31 | 2027-12-31 |
| DisplayHow the image is produced. "No display" is a real answer in this category, not a gap. | Waveguide | Waveguide |
| Eyes coveredOne eye or both. Over a full shift this is a real difference in fatigue. | Both eyes | Both eyes |
| Resolution per eyePixels delivered to each eye. Higher means small text and labels stay readable. | ≈1280x960*VRcompare | 1440x1760 |
| Refresh rateImages per second. Higher feels smoother and helps people prone to motion discomfort. | ≈122 Hz*VRcompare | ≈120 Hz*Magic Leap 2 Detailed Spec Sheet (PDF, distributor-hosted) |
| Field of viewHow much of your view the image covers. Bigger is more immersive, not automatically better for work. | ≈50° diagonal*VRcompare estimate | 70° diagonal |
| Pixel densityPixels per degree of view — the honest way to compare sharpness between a headset and a pair of glasses. Higher is sharper. Only shown where somebody published it; we never compute it ourselves. | Pixel density: not published by the manufacturer | Pixel density: not published by the manufacturer |
| BrightnessIn a plant or outdoors, brightness decides legibility more than resolution does. | Brightness: not published by the manufacturer | ≈2000 nits*VRcompare |
| WeightThe device alone. Anything above roughly 150 g is felt within an hour. | ≈316 g*VRcompare | 260 g |
| Visible field of viewWhat the eye actually sees through the optics, by axis. | 40° horizontal · 30° vertical · 50° diagonal | Visible field of view: not published by the manufacturer |
| Rendered field of viewWhat the device draws. Never smaller than what you see. | Rendered field of view: not published by the manufacturer | Rendered field of view: not published by the manufacturer |
| Binocular overlapWhere both eyes see the same thing — this is where depth happens. | Binocular overlap: not published by the manufacturer | Binocular overlap: not published by the manufacturer |
| Average pixel densityPixels per degree across the image. More is sharper. | Average pixel density: not published by the manufacturer | Average pixel density: not published by the manufacturer |
| Peak pixel densityThe sharpest point, usually the centre of the lens. | Peak pixel density: not published by the manufacturer | Peak pixel density: not published by the manufacturer |
| Subpixel layoutHow the coloured dots sit in a pixel. A stripe looks cleaner on text than PenTile. | Subpixel layout: not published by the manufacturer | Subpixel layout: not published by the manufacturer |
| Foveated renderingWhether detail is dropped outside where you look, to save performance. | Foveated rendering: not published by the manufacturer | Foveated rendering: not published by the manufacturer |
| Vision correctionHow somebody who wears glasses uses it, and in what dioptre range. | Vision correction: not published by the manufacturer | Official lens inserts · -10 to 5 D |
| DimensionsWidth, height and depth of the device itself. | Dimensions: not published by the manufacturer | Dimensions: not published by the manufacturer |
| Weight with strapWith the strap it ships with — the weight you actually wear. | Weight with strap: not published by the manufacturer | Weight with strap: not published by the manufacturer |
| Head strapSoft band or a rigid one. Decides how the weight sits over an hour. | Rigid strap with a dial | Rigid strap with a dial |
| MaterialsShell and facial interface, in the vendor’s words. | plastic · foam | Plastic |
| ColoursColours the manufacturer sells it in. | Grey | Grey |
| Haptics in the headsetVibration in the headset itself, not in the controllers. | No | No |
| TrackingWhether content stays put in the room (6DoF), follows head rotation (3DoF) or is simply fixed. | 6DoF (position in space) | 6DoF (position in space) |
| Tracking camerasHow many cameras it uses to know where it is. | Tracking cameras: not published by the manufacturer | 3 |
| Tracking rateHow often position is recalculated. Higher is steadier. | Tracking rate: not published by the manufacturer | Tracking rate: not published by the manufacturer |
| Depth sensorA dedicated sensor for distance, on top of the cameras. | Depth sensor: not published by the manufacturer | Yes |
| Hand trackingWhether bare hands can operate the interface, without a controller or the phone. | Yes | Yes |
| Hand tracking detailHow much of the hand reaches the application: a pinch, or a jointed hand pose a scenario can read. | Full hand skeleton | Full hand skeleton |
| Hands and controllers at onceWhether a controller in one hand and a bare hand in the other work at the same time. | Hands and controllers at once: not published by the manufacturer | Hands and controllers at once: not published by the manufacturer |
| Hand tracking SDKThe SDK an application reaches the hand tracking through, and any named modes that change how it behaves. | Hand tracking SDK: not published by the manufacturer | Hand tracking SDK: not published by the manufacturer |
| Face trackingTracks facial expression — for avatars, not for training scenarios. | Face tracking: not published by the manufacturer | No |
| Body trackingHow much of the body the device knows about. Estimated legs are not tracked legs. | Body tracking: not published by the manufacturer | Body tracking: not published by the manufacturer |
| Controllers includedWhether motion controllers come in the box, or are bought separately — or the device is hands-only. | Yes | Yes |
| Controller modelThe manufacturer's own name for the controller. | Magic Leap Controller | Magic Leap 2 Controller |
| Controller tracking6DoF controllers track position and rotation, so a hand reaches in space; 3DoF only point. “None” means hands only. | 6DoF | 6DoF |
| Controller tracking methodWhat establishes where the controller is. It decides whether the controller keeps working behind your back or down at your hip — “6DoF” on its own does not say. | Magnetic field | Cameras on the controller |
| Other controllersWhether the headset works with controllers other than the ones it ships with. | Other controllers from the same maker — Magic Leap Mobile App on an iOS or Android phone, as a second 3DoF Control | This headset’s own controllers only |
| Finger tracking on the controllerWhether the controller senses individual fingers, and how. | Finger tracking on the controller: not published by the manufacturer | Finger tracking on the controller: not published by the manufacturer |
| Controller inputsButtons, sticks and triggers the controller offers. | Controller inputs: not published by the manufacturer | Touchpad · trigger · bumper · menu button · home button |
| Controller hapticsVibration feedback in the hand. | Yes | Yes |
| Controller powerDisposable cells or a built-in rechargeable — a fleet question. | Controller power: not published by the manufacturer | Built-in rechargeable |
| Controller weightOne controller, as the vendor weighs it. | Controller weight: not published by the manufacturer | 140 g |
| Other input methodsWays in beyond controllers and bare hands. | Other input methods: not published by the manufacturer | Voice |
| Where it computesWhether the device computes on its own or depends on a phone, PC or belt-worn puck. | Driven by a belt-worn puck | Driven by a belt-worn puck |
| ProcessorThe processor and memory inside. This is what the performance grade above is derived from. | NVIDIA Tegra X2 (Parker SoC) | ≈AMD Quad-core Zen2 x86 CPU*VRcompare |
| CPUCores and clocks, where the manufacturer publishes them. | 2x Denver 2.0 + 4x ARM Cortex-A57 | Quad-core Zen2 (4 x 3.92 GHz, 8 threads) |
| GPUThe graphics part of the chip. | NVIDIA Pascal, 256 CUDA cores | AMD GFX10.2: 1SE 1SA 4 WGP |
| MemoryRAM on the device. Together with the chipset it sets how complex a scene can be. | 8 GB | 16 GB |
| PlatformThe platform an application is built against — and what an MDM has to speak. | Lumin OS | Magic Leap OS (AOSP) Android 10 |
| StorageOn-board space for downloaded courses. More matters for an offline library in several languages. | Storage: not published by the manufacturer | Storage: not published by the manufacturer |
| SD card slotWhether storage can be added with a card. | SD card slot: not published by the manufacturer | SD card slot: not published by the manufacturer |
| BatteryRuntime as the manufacturer states it. A hot-swappable pack matters more than the figure. | 180 min | 210 min |
| Battery capacity (Wh)Energy in the cell. Bigger is longer, given the same power draw. | Battery capacity (Wh): not published by the manufacturer | Battery capacity (Wh): not published by the manufacturer |
| Battery capacity (mAh)The same fact in the other unit vendors use. We never convert between the two. | Battery capacity (mAh): not published by the manufacturer | Battery capacity (mAh): not published by the manufacturer |
| Charge timeFrom empty to full. Rarely stated by the manufacturer, so it usually comes from whoever timed it. | Charge time: not published by the manufacturer | 180 min |
| ChargerPower of the supplied charger — it decides the charge time above. | Charger: not published by the manufacturer | Charger: not published by the manufacturer |
| External batteryBattery on a cable or a puck instead of on the head. | External battery: not published by the manufacturer | Yes |
| Charging caseExtra runtime carried in the case the glasses ship with. | Charging case: not published by the manufacturer | Charging case: not published by the manufacturer |
| Hot-swap batteryWhether a charged battery can replace the empty one without ending the session. This is what decides if a headset can hold an all-day training station. | Hot-swap battery: not published by the manufacturer | Hot-swap battery: not published by the manufacturer |
| Third-party appsHow a third party's application reaches the device: an ordinary Android package, a vendor-only SDK, or not at all. | Vendor SDK only | Ordinary Android app (APK) |
| Runs Skillsive appsOur own statement about whether Skillsive training applications run on it today. | No | No |
| Suitable for trainingOur judgement on whether it survives real deployment: comfort, hygiene, fleet management. | No | Suitable for training: not published by the manufacturer |
| Kiosk mode / MDMVendor tooling to enrol devices, push an app and lock the headset to it. Practically required above ten devices. | No | Yes |
| Business tierAn official business tier, so devices belong to the company rather than to employees’ personal accounts. | No | Business tier: not published by the manufacturer |
| CertificationsIndustrial certifications stated on the manufacturer's own page. Empty means unconfirmed, never "probably fine". | Certifications: not published by the manufacturer | FCC · CE · ANSI/CAN/UL 8400 · IEC 62471 · IEC 60825 · ISO 10993 |
| PortsPhysical sockets and charging contacts. | USB Type-C | USB Type-C |
| Wi-FiThe Wi-Fi generation — it decides how well wireless streaming holds up. | 5 | 6 (802.11 a/b/g/n/ac/ax) |
| BluetoothFor controllers, headphones and accessories. | 4.2 | 5.1 |
| Video from a computerWhether a computer can drive it, over a cable or wirelessly. | Video from a computer: not published by the manufacturer | Video from a computer: not published by the manufacturer |
| AudioBuilt-in sound. Matters most where somebody has to hear a remote expert while working. | Built-in speakers | Built-in speakers |
| MicrophoneBuilt-in microphone — the other half of talking to a remote expert. | Yes | Yes |
| MicrophonesHow many. An array picks a voice out of a noisy hall; one does not. | Microphones: not published by the manufacturer | 4 |
| CameraA camera is what makes a remote expert call and photo documentation possible. | Yes | Yes |
| Price bandThe band it sits in, rather than a price that goes stale. | Enterprise | Enterprise |
| Launch priceThe manufacturer's launch price. Fixed history, so it never goes stale. | $2295 | Launch price: not published by the manufacturer |
| Price by configurationWhat each configuration cost at launch. | Price by configuration: not published by the manufacturer | Price by configuration: not published by the manufacturer |
| Price todayApproximate current net price in USD — a general figure, since these barely differ between regions once tax is stripped out. A snapshot from the date shown, not an offer. | Price today: not published by the manufacturer | Price today: not published by the manufacturer |
“—” means the manufacturer does not publish that figure. “≈” with an asterisk means it was measured by somebody else — follow the asterisk for who. Prices are an indicative net figure in USD, a snapshot from the date on each model page, not an offer.
Product and company names used here are the trademarks of their respective owners and appear only to identify the hardware being compared. We are not affiliated with, sponsored by or endorsed by any of these manufacturers. Illustrations are our own.