The clip looks almost too perfectly engineered for the internet: an Evangelion reaches toward a red-eyed Angel, a hand pulls a trigger inside Tokyo-3, Misato reaches directly towards the player and—because this is still Evangelion—someone calmly trims a hedge while Rei watches from a safe distance.
It has prompted an obvious question across social media: is EVANGELION: Δ CROSS REFLECTIONS a real VR game, or are we watching an unusually convincing AI concept trailer?
The short answer is reassuring, with one necessary asterisk. The game is real. The licence is real. Public demos are real. The circulating video was released by the project’s official accounts. What has not been established is that every second of the polished teaser represents raw, unedited footage captured directly from a retail headset.
That distinction matters, because Cross Reflections might have the ingredients to become the anime game that finally persuades people to buy a VR headset.
Yes, Evangelion: Cross Reflections is a real licensed game
Pixelity Inc. first announced its licensing agreement with Studio Khara in February 2025. The Korean XR developer later revealed the full title and outlined something much more ambitious than a short location-based attraction.
Cross Reflections is planned as the first chapter in a trilogy covering the 26 episodes of the original Neon Genesis Evangelion television series. Part one is built around episodes one through 11, but players will not simply replay Shinji’s story. Pixelity casts us as an original NERV officer cadet whose path crosses established characters, locations and events.
The developer’s official project description names two main experiences. Story Mode puts the player inside Tokyo-3 and NERV headquarters, while Battle Mode moves the action into a NERV simulation environment. VR and mixed reality are both part of the plan.
The original 2026 window has slipped. Pixelity now says the first game is arriving in 2027, although there is no exact date, price or confirmed list of retail headsets. “Various VR devices” is the most specific official platform language available.
Is the viral Evangelion VR clip AI?
We found no credible evidence that the viral teaser is an AI-generated fake.
The 42-second video was posted by the official Japanese Cross Reflections account on September 29, after being screened at Pixelity’s San Francisco hands-on event. It carries the project’s Khara and Pixelity copyright marks and a clear warning that the pre-release footage may differ from the final product.
More importantly, Cross Reflections has existed outside a video file. Pixelity put an early build in front of players at Evangelion’s 30th anniversary event in Yokohama, showed its controller-free hand-tracking system at BitSummit and invited roughly 80 people to try Story and Battle modes in San Francisco.
A San Francisco attendee later described playing on Meta Quest 3 hardware with no hand controllers. The demonstration used hand tracking, haptic wrist devices and a vest. Their session included a NERV story mission supporting Unit-01 during the battle with Sachiel, plus a combat simulation with punches, a pallet rifle, an AT Field and a sync-rate system. A separate Japanese hands-on report also documented a Quest 3 demo using hand tracking and a haptic suit.
Those accounts do not prove that the trailer is an untouched recording from a headset. They prove that a functioning game exists and that several of the ideas shown in the trailer are playable.
Official footage is not the same as raw gameplay
This is where healthy scepticism belongs.
As the published teaser makes clear, the footage is tightly edited, has no uninterrupted play session, no visible performance readout and no confirmed target headset. It may contain engine footage, staged camera angles, post-processing or sequences rendered above the quality a standalone headset will deliver. The disclaimer exists for a reason.
Hands-on impressions provide a more grounded picture. The cel-shaded art apparently preserves the anime’s look well, while resolution and overall fidelity were described as acceptable rather than miraculous. That sounds believable for a Quest 3 development build. It also explains why the trailer can feel cleaner than footage recorded through headset lenses.
So our verdict is precise: the video is authentic official marketing for a real, playable project; it is not yet proof of final consumer graphics. Calling it AI because it looks unusually polished skips over a substantial trail of public demonstrations. Calling it literal final gameplay would skip over the same trailer’s warning.
The strange little moments may be its smartest idea
The most encouraging shots are not necessarily the Angel battles. They are the mundane ones.
The teaser shows the player handling glasses, operating a phone, repairing a vehicle, eating pie and performing the now-famous hedge trim. It looks less like a 40-minute cockpit ride and more like an attempt to let us inhabit the spaces between Evangelion’s catastrophes.
That is an astute reading of the source material. The Evas feel impossibly large because Evangelion first teaches us what a school corridor, railway platform, apartment kitchen and awkward silence feel like. VR can make those ordinary spaces personal before it asks us to look up at a creature the size of a building.
It is the same design lesson that makes a native experience such as How to God work: embodiment matters more than simply placing a familiar licence around the player.
Could this really be VR’s killer app?
It has a better chance than most licensed VR projects.
Evangelion gives virtual reality three things the medium desperately needs: a globally recognised audience, a fantasy that depends on physical scale and a cockpit people have wanted to enter for 30 years. Hand tracking removes the small psychological barrier of learning unfamiliar controllers. Reaching, shielding, aiming and forming an AT Field with your own hands are instantly understandable.
The visual language helps too. Cel shading can look intentional on modest hardware, where a chase for photorealism quickly exposes every compromise. Pixelity can preserve the series’ graphic identity while spending its performance budget on scale, animation and responsiveness.
But a killer app must do more than delight existing headset owners. It must make people buy hardware. That will depend on questions Pixelity has not answered: which headsets are supported, whether the three chapters are sold separately, how long each chapter lasts, whether controllers are optional, how much mixed reality is required and what home players lose without the haptic equipment used at demonstrations.
Platform support could decide everything. The Quest 3 demos show that Pixelity has a working standalone build, but the company has not announced a Quest release. A broad launch across Quest, PC VR and PS VR2 would give the game reach; a fragmented or expensive release would narrow it immediately. Anyone weighing the next generation of hardware already faces enough uncertainty, as our comparisons of Steam Frame and Quest 4 and Meta’s new VR glasses have made painfully clear.
There is also a scale warning hiding inside the word “trilogy.” Pixelity is a specialist XR studio, not a publisher with unlimited money. Splitting 26 episodes across three releases makes production possible, but it creates three chances for price, pacing or delays to exhaust the audience.
What we still need to see before 2027
The next reveal should be boring in the best possible way: ten uninterrupted minutes recorded from the player’s perspective.
Show us the menus, locomotion, comfort settings, hand-tracking failures, loading transitions and a complete battle. Confirm the headsets. Explain whether the Quest 3 demonstration represents the baseline version. Tell us how haptics scale down to an ordinary living room.
Pixelity does not need another impossible trailer. It already has the attention.
Cross Reflections is real, and the viral clip is real promotional footage from that project. The game’s killer-app status remains a prediction, but the trailer has already completed the first test: it made Evangelion fans who never cared about VR imagine buying a headset.
For this medium, that may be the most valuable signal of all.
About the author
The conversation
Comments are closed.