TikTok parent ByteDance is reportedly preparing an AI model that could let you create real-time spatial videos similar to Google Genie.

According to a Bloomberg report, ByteDance is readying an AI model dedicated to creating real-time spatial videos, which the company hopes will act as a “flywheel” to its various companies, including video streaming giant TikTok and its XR hardware subsidiary, Pico Interactive.

Citing people familiar with the matter, the report maintains founder Zhang Yiming is personally overseeing the development as it nears launch as soon as next month.

The AI model is thought to work similar to Google’s Genie, which not only generates interactive virtual environments in real-time, but also can simulate physical environments for AI agents.

Notably, Zhang reportedly sees the tech as a major component linking the company’s existing AI models and cloud computing resources with its content platforms and XR headset subsidiary, Pico Interactive.

The report maintains the model is being built on Seedance, the company’s AI model for generating cinematic videos.

There’s no word on timing outside of that “as soon as next month” window, however Pico delayed the reveal of its upcoming Apple Vision Pro competitor, Space Pro, due to “spend[ing] more time on a major upgrade to the software experience.”

It’s unclear whether ByteDance’s real-time spatial video model is the main reasoning behind the delay, or something else.

Speculatively speaking, Pico may also be facing issues as Valve upcoming Steam Frame standalone headset, which is primarily due to the ongoing global component shortage—seeing RAM, storage, and reportedly also now Qualcomm SoCs prices skyrocket.

The company says it will unveil Pico Space Pro sometime in Q4, so we’re hoping to learn more soon, as Q4 is due to officially start October 1st.

Pico’s upcoming Apple Vision Pro competitor, Space Pro, was initially meant to debut earlier this month, although the ByteDance subsidiary cancelled the event and postponed its reveal and release date. Now, the company’s next flagship standalone XR headset just cleared the FCC.

Previously known as ‘Project Swan’, Pico Space Pro has now surfaced in FCC filings, which is typically one of the last hurdles before a company can commercialize electronics in the US. The news was first reported by UploadVR.

At the event, which was supposed to take place on September 2nd in Beijing, we were hoping to see a more complete spec sheet, as well as learn the price and release date. That said, much is already known about the device, which is referred to in the filings as a ‘Mixed Reality Computer’.

Space Pro is set to include microOLED displays boasting 4,000 pixels per inch (PPI), an average ~40 pixels per degree (PPD), and a dual-chip architecture: a custom XR silicon for perception and imaging (claimed ~12 ms perception latency) and separate flagship SoC with “more than 2× CPU and GPU performance vs XR2 Gen 2.”

Pico hasn’t specified exactly when it plans on rescheduling Space Pro’s reveal, although the company says we should expect launch sometime in Q4 2026.

Although the company maintains the delay was to “spend more time on a major upgrade to the software experience,” the timing suggests Space Pro may also be affected by the global component shortage, which has not only seen RAM and storage prices skyrocket, reportedly now Qualcomm SoCs too.

As it is, Pico seems to be taking a slower, more deliberate approach to launch though. When it cancelled the device’s reveal (just one week before the actual event), the company announced it was instead expanding its ‘Global Early Access Program’ closed beta to a number of hand-picked testers, which includes sending pre-production units and software to applicants.

Pico Space Pro isn’t only set to compete against Apple Vision Pro, but also Samsung’s Galaxy XR headset—the first to run Google’s Android XR processing system—in addition to Valve’s Steam Frame, which still doesn’t have a price or release.

China-based projector manufacturer XGIMI launched a crowdfunding campaign in June for its first smart glasses, MemoMind One, which is banking on consumer interest in display-clad smart glasses that don’t contain cameras. Now the campaign has reached $1 million, giving the segment anther early vote of confidence.

The News

MemoMind One quickly surpassed its lowball fund goal of $80,000 HKD (~$10,000 USD) when the Kickstarter launched on June 28th, going on to secure $500,000 in less than a week after launch. Now the project just topped $1 million with 20 days left in the campaign, which closes on August 28th.

If you haven’t followed its progress, MemoMind One was first shown at CES 2026 in Las Vegas in January, pitching a pair of smart glasses that don’t include any sort of camera, instead sporting dual green micro-LED displays (2,000 nits), integrated audio, and a host of AI-assisted actions, such as real-time translation, note taking, contextual assistance, and navigation.

Notably, MemoMind One is positioning itself as one of the least expensive display smart glasses on the market, with pricing starting at $400 for the Standard Edition of all stylings: Nomad, Gotham and Archive. There is also the option to add prescription lenses ($500) and a Custom Edition with multiple colorways starting at $450, all of which are expected to ship in August 2026.

XGIMI hasn’t released any sort of stretch goals, as it appears the campaign is more oriented toward pre-orders. You can learn more about it on MemoMind One’s Kickstarter here.

My Take

This is a bit of a casual rehash of my original take when MemoMind One reached the $500,000 mark, but I’ll save you click with this (slightly) condensed version.

If MemoMind One’s campaign is any indication, not having an onboard camera is a much stronger marketing position than you might think, partly due to blowback Meta is getting from its own display-less camera smart glasses right now.

Until recently, Meta hasn’t properly kept creepers from drilling out capture LEDs, or otherwise vaunting the system by other means. It’s too early to tell since there are no major competitors yet, but Meta may have ‘poisoned the well’ somewhat.

While MemoMind isn’t the first to claim camera-less display smart glasses as a unique marketing proposition, it is doing it at a cheaper price than the main competition: Even Realities G2, which are priced at $600 before adding prescription lenses.

That said, I’m not certain ‘display only’ smart glasses will ever be truly massive; cameras and displays are difficult to run together in today’s power and compute constraints, so it seems more like a temporary division that’s being fashioned into a marketable segment. I’ll be putting those assumptions to the test when I go in for a long-term hands-on with MemoMind One, so check back soon.

VITURE today launched its latest AR glasses, VITURE Pro 2, which the company calls its thinnest, lightest and comfortable yet. Alongside it, Viture also unveiled a new, more pocketable USB-C dock which lets you connect to a wider range of devices, such as Nintendo Switch 2.

Viture Pro 2 – $300

Like many of the company’s AR glasses, Viture Pro 2 is mostly meant for traditional gaming and computing, albeit on a virtual screen projected in front of you thanks to its see-through AR birdbath optics.

And while it may look similar to the original 2024 Viture Pro from the outside, the company says it’s thinner, lighter, and brighter. Its optical stack has also been re-tuned optical stack, delivering better clarity overall, the company says.

Dubbed ‘UltraClarity 3.0’, the Viture says it’s tightened four key visual factors over its predecessor, including field curvature, image sharpness across the full 50-degree field of view (FOV), geometric distortion across the entire diopter range, and image stability as the pupil moves off-center.

“Sharpness holds not just at the center but into every corner,” the company says. “Horizontal and vertical resolve equally. Straight lines stay straight at any prescription. And the image stays put wherever the eyes wander, instead of swimming as the pupil moves off-axis.”

It’s also impressively thin and light, weighing just 63g, which the company notes is also the thinnest frame the company has ever built, making it 20% lighter and 16% thinner than the original Viture Pro.

Note: what you don’t see in Viture Pro 2’s promo photos is the required USB-C tether, as it’s not a standalone device, requiring you to attach the included cable from the left temple tip of the glasses right behind the ear. Equally important to note is that Viture Pro 2 doesn’t support 6DOF spatial tracking, as it lacks onboard camera sensors, meaning you can’t pin a display to a specific point in space.

Under the hood, or rather sandwiched on top of its birdbath-style AR optics, is a pair of Sony micro-OLED displays delivering the same 1,920 x 1,080 per-eye resolution and 120Hz refresh rate, albeit with a higher perceived brightness of up 1,600 nits over the original’s 1,000 perceived nits.

It also boasts a built-in diopter adjustment with independent dials per-eye, covering 0 to −5.00D, a slightly higher FOV of 50 degrees (originally 46 degrees), and three included nose pads for seven adjustment positions.

You can find Viture Pro 2 available today for $300 across Viture and Amazon.com, which includes a free lens shade in the box for better immersion.

Mobile Dock Mini – $100

As a part of the company’s five-year anniversary, Viture also unveiled a new, smaller mobile dock called Mobile Dock Mini ($100), which is now available for pre-order.

While Viture Pro 2 can already connect to a variety of USB-C devices natively, the modular, pocket-sized dock expands that to more devices —notably Nintendo Switch 2—while doubling as a USB-C hub and extended battery.

In all, it’s essentially a more compact version of the company’s original Viture Pro Mobile Dock ($100), featuring a slight smaller at 10,000 mAh battery than the original’s 13,000 mAh. The new Mobile Dock Mini is however 34% smaller and 30% lighter, making it a more pocketable way to charge a phone, Nintendo Switch 2 while powering any of Viture’s glasses models at the same time.

It works in two different modes: ‘Hug Mode’, which “hugs” the back of a Switch 2 as a single unit via a magnetic mount, and ‘Solo Mode’, which allows it to become a compact standalone USB-C hub.

You can find this and the whole gamut of Viture glasses and accessories over on Viture’s website.

Snap announced it’s hosting an event dedicated to Specs, the company’s next-gen AR glasses, on September 16th in Los Angeles, California.

At the event, Snap co-founder and CEO Evan Spiegel is said to share “the first in-depth look at SPECS, demonstrating how they bring AI assistance, work tools, entertainment, and shared experiences into the world around us, helping people create, connect, learn, and get things done while staying present,” the company says.

While the company characterizes it as a “launch” event for Specs, the device’s pre-order page maintains it’s still slated to ship sometime this fall across the US, UK and France, suggesting we’ll get a more precise ship date during the event.

Snap launched pre-orders for Specs in June, priced at $2,195. Similar to the company’s fifth-gen Specs released in 2024, the new device don’t require an external compute puck or tethered connection to external devices, as the glasses pack in two Qualcomm Snapdragon processors—one dedicated to computer vision and another for running AR experiences, or “Lenses,” as the company calls them.

In the meantime, you can sign up to watch the Specs event live here, which takes place on September 16th at 4:00 PM PT (local time here). Attendees of the physical event will have an opportunity to be some of the first to go hands-on.