Meta Quest 4 may be delayed until 2027 — but a lightweight alternative could take its place next year

The Meta Quest 4 is being pushed back until 2027, according to a recent leak, but we may be seeing an all-new lightweight VR device in 2026 instead — and it comes with an external puck.
Meta appears to be shifting focus for its upcoming VR headset plans, with notable leakers Brad Lynch and Luna on X stating that the company may switch the release of the Quest 4 from next year to 2027.
This is apparently due to two planned prototype headsets, codenamed "Pismo Low" (a budget model similar to the Quest 3S) and "Pismo High" (a high-end headset akin to the Quest 3), being canceled.
According to the leaks, a "high-end device" is looking to arrive in 2026 instead, and it's set to be a lightweight headset that runs Horizon OS with an external compute puck, codenamed "Puffin."
Yes, I too have heard whispers that Meta may swap the release schedule and debut their new high-end device in 2026, rather than Quest 4. Quest 4 would then ship a year later.June 2, 2025
We've heard that Meta was developing a Quest Pro "successor" instead of a Quest Pro 2, and it appears Puffin could fit into that slot. According to RoadtoVR, this ultralight device would resemble VR goggles rather than a headset, and will be less focused on gaming than Meta's usual headsets.
Instead, Puffin is rumored to be all about virtual screens, offering users a multi-screen setup for entertainment and productivity purposes. This isn't unlike some of the best AR glasses, including the upcoming Viture XR smart glasses.
However, it stands out as the Horizon OS-based device is tipped to come with that compute puck — small enough to fit in a jacket pocket. This would be tethered to the device and deliver processing power, making it lighter to wear. It isn't unlike the Apple Vision Pro, although that headset's external device is a battery.
Get instant access to breaking news, the hottest reviews, great deals and helpful tips.
Rumors also indicate Meta's Puffin device won't arrive with any controllers, instead using hand-tracking to navigate around menus. That's what we've seen in the Quest 3 and Vision Pro already, but it could mean it won't offer any gaming-centric features.
Interestingly, Lynch indicates Valve's Deckard headset will be available in 2026, with previous details suggesting it could launch in Fall 2025. Along with an upcoming Asus ROG VR headset (set to be the first third-party headset to use Horizon OS) in the works, there are a few VR devices coming our way.
While the Quest 4 may not arrive until 2027 now, Meta's upcoming Puffin is still looking to be at the forefront of next-gen VR devices — and Apple may be scrambling to beat it to the punch.
According to 9to5Mac, the Cupertino tech giant wants to launch its Apple Glasses ahead of Meta's offerings, but with Meta's changed plans, we could see its AR glasses get pushed up. What's more, rumor has it that the next Apple Vision Pro headset could drop sooner than we think.
So 2026 is shaping up to be a big year for VR and AR devices, and while we'll have to wait and see what Meta ends up delivering, along with other headsets from major competitors, it's clear we have a lot to look forward to.
That's not all we have coming our way, as we may see Meta's $1,000 smart glasses with a built-in screen arriving this year.

Try Our AI Features

Explore what Daily8 AI can do for you:

Learn with AIFact CheckingText to SpeechAI Summary

Related Articles

Yahoo

22 minutes ago

• Yahoo

Steam's store now lets you search for games by accessibility features like 'narrated game menus' and 'adjustable difficulty', and more than 5,000 games have already added their accessibility details to the database

When you buy through links on our articles, Future and its syndication partners may earn a commission. Back in April, Valve announced the Steam store would start highlighting accessibility features on games' store pages. And now, it has. Not only can you see them in the sidebar beneath the controller support section for selected games, you can also search the store by individual accessibility features. "This update comes after gathering valuable feedback from developers as well as players with disabilities," Valve says in its blog post announcing the news, "and over 5,000 applications have added details about their accessibility support (with more developers updating their games each day)." On Steam's search page you can select "narrow by accessibility feature" to bring up a bunch of toggles like "playable without timed input" and "color alternatives". If you know anyone who has ever bought a puzzle game and had to refund it because the puzzles demand you can tell colors apart but won't let you tweak what those colors are, you'll know why the latter's important. Some accessibility features, like "save anytime" and "adjustable text size" are good for everyone to know in advance—they're just nice things to have. A variety of games, mostly newer ones, have already added accessibility features to their store page. Baldur's Gate 3, for instance, boasts 10 of them, from "adjustable difficulty" to "playable without timed input". Here's a full list of the options, with explanations of exactly what they mean. Gameplay Adjustable Difficulty: Players can adjust gameplay difficulty. Save Anytime: Players can save gameplay using both automatic and manual saves. Saving can be performed at any point in the game. Visual Adjustable Text Size: Players can adjust text size. (In-game text, menu text, character dialog text, subtitle text) Subtitle Options: Players have options to customize the display of subtitles for all spoken content and essential audio information. Color Alternatives: Gameplay doesn't rely on colors to communicate important information, or players have an option to adjust colors used for distinguishing information. Camera Comfort: Players have an option to adjust or disable uncomfortable camera movement such as screen shaking, camera bob, or motion blur, or the game doesn't feature these effects. Audio Custom Volume Controls: Players can adjust the volume of the audio. Different types of audio can be muted independently from each other. Narrated Game Menus: Players can listen to game menus with narrated audio. Stereo Sound: Players can identify how far to the left or right sounds are coming from. Surround Sound: Players can identify how far in any direction sounds are coming from. Input Keyboard Only Option: Players can play the game with just a keyboard and no other additional input mechanisms such as a mouse or controller. Mouse Only Option: Players can play the game with just a mouse and no other additional input mechanisms such as a keyboard or controller. Touch Only Option: Players can play the game with just touch controls and no other additional input mechanisms such as a mouse, keyboard or controller. Playable without Timed Input: Players have an option to adjust gameplay to not require precisely timed button presses ("quick time events") or gameplay does not require such button presses. Text-to-speech: Text chat can be narrated out loud in real time. Speech-to-text communications: Voice chat can be read as a text transcript in real time. Baldur's Gate 3 romance: Who to pursueBaldur's Gate 3 multiplayer: How co-op worksBaldur's Gate 3 endings: For better or worseBaldur's Gate 3 multiclass builds: Coolest combosBest RPGs: The greatest you can play now

Forbes

an hour ago

• Forbes

Meta Invests $14 Billion In Scale AI To Strengthen Model Training

Big Data Meta's $14.3 billion investment in Scale AI represents the social media giant's most significant move to secure high-quality training data for artificial intelligence models. The deal gives Meta a 49% stake in the data labeling startup while bringing Scale AI founder Alexandr Wang into Meta's leadership to head a new superintelligence research lab. This acquisition addresses Meta's most pressing challenge in the AI race: access to the specialized datasets required to train competitive large language models. While competitors like OpenAI lead the global AI market share through ChatGPT, Meta's recent Llama 4 models received lukewarm reception from users who reported poor performance in coding tasks and generic responses compared to smaller rivals. The Data Foundation Problem Scale AI operates a global workforce of contractors across Kenya, the Philippines and Venezuela who manually label images, text and video for machine learning applications. The data labeling process involves human annotators identifying objects in images, transcribing audio or categorizing text to create the training datasets that teach AI models to recognize patterns. For autonomous vehicle applications, this includes labeling 3D point clouds from lidar sensors and marking objects across video frames. In natural language processing, workers rate the quality of AI responses and provide feedback through reinforcement learning techniques that incorporate human feedback. Meta's investment secures privileged access to these data preparation services while its competitors face potential service restrictions. Google paused multiple Scale AI projects within hours of the announcement of the Meta deal. OpenAI confirmed it was already winding down its Scale AI relationship, and Elon Musk's xAI halted some projects as well. Market Disruption and Competitive Response Scale AI differentiates itself through its integrated platform capabilities, which combine data labeling, model evaluation, and synthetic data generation. The company's workforce includes highly educated and skilled contractors with PhDs and master's degrees. This expertise proves critical for complex domains like healthcare, finance and legal services that require a nuanced understanding beyond basic image recognition. The Meta investment creates immediate market consolidation as Scale AI's major clients seek alternative providers. This shift benefits competitors like iMerit, which leverages domain expertise in healthcare and geospatial applications, and automated labeling platforms such as Snorkel AI that reduce dependence on human annotators. Technical Integration and Capabilities Wang will lead Meta's new superintelligence lab focused on developing artificial general intelligence. The 28-year-old MIT dropout previously worked at high-frequency trading firm Hudson River Trading before founding Scale AI in 2016. His team of approximately 50 researchers will join Meta's existing AI workforce as the company plans to spend a significant amount on AI infrastructure in 2025. The integration provides Meta with several technical advantages. Scale AI's data engine processes multiple modalities, including text, images, video and audio, through both automated systems and human oversight. The platform features quality assurance mechanisms that utilize statistical sampling to identify edge cases, resulting in a substantial reduction in revision cycles. Meta's access to Scale AI's government contracts also expands its reach into defense applications. Wang's connections in Washington could help Meta secure federal AI projects, diversifying beyond its consumer-focused social media platforms. Strategic Implications for Enterprise AI The deal structure avoids traditional acquisition scrutiny by maintaining Scale AI as an independent entity while giving Meta operational control. This approach mirrors Microsoft's OpenAI investment and Amazon's Anthropic backing, allowing tech giants to access AI capabilities without triggering antitrust reviews. For enterprise technology leaders, Meta's move signals the critical importance of data quality in AI implementations. Nearly all business leaders report encountering AI-related data quality issues, with problems including duplicate records, privacy constraints and inefficient integration hampering deployment goals. The Meta-Scale AI partnership demonstrates that even well-funded companies struggle with the foundational data challenges that determine AI success. The investment also highlights the growing strategic value of specialized AI infrastructure. While enterprises often focus on model selection and deployment, the quality and diversity of training data ultimately determines system performance. Companies that secure reliable data labeling capabilities gain sustainable competitive advantages in AI applications. Meta's willingness to pay $14.3 billion for a data services company reflects the market reality that high-quality training data has become the primary constraint on AI development. As the global AI market continues expanding, access to specialized data preparation capabilities will increasingly separate successful AI implementations from failed projects. The deal positions Meta to compete more effectively against OpenAI and Google by addressing its most significant disadvantage: limited access to the diverse, high-quality datasets required for advanced AI model training. Whether this investment translates into improved AI products remains dependent on Meta's ability to integrate Scale AI's capabilities with its existing research and development efforts.

Wall Street Journal

5 hours ago

• Wall Street Journal

Zuckerberg Leads AI Recruitment Blitz Armed With $100 Million Pay Packages

The smartest AI researchers and engineers have spent the past few months getting hit up by one of the richest men in the world. Mark Zuckerberg is spending his days firing off emails and WhatsApp messages to the sharpest minds in artificial intelligence in a frenzied effort to play catch-up. He has personally reached out to hundreds of researchers, scientists, infrastructure engineers, product stars and entrepreneurs to try to get them to join a new Superintelligence lab he's putting together.