VR game development: a practical guide for studios

Making a VR game starts with a clear hardware target, an engine that supports OpenXR, a comfort-first interaction model, and a performance budget that holds a steady 90 frames per second on the chosen device. This guide walks through scoping, engine and SDK choice, locomotion, performance optimisation, store submission, and team setup for standalone and PCVR projects.

Scope VR game development before touching an engine

VR scope decisions made in pre-production define almost every later cost. The first call is session length and play space: seated, room-scale, or arena. The second is the player count and whether the game is solo, co-op, or social. The third is the target hardware tier, which sets the polygon and draw-call ceiling. A standalone Quest 3 title runs on a Snapdragon XR2 Gen 2 with no PC tether, so it lives inside a tight mobile-class budget. A PCVR title built for Steam can lean on a desktop GPU and afford more shader complexity. PSVR2 sits in between, with eye tracking and dynamic foveated rendering as built-in tools. Pin all three answers before the first prototype.

Typical early deliverables before production scales:

VR game development hardware: standalone vs PCVR vs PSVR2

Hardware choice is the largest single driver of scope. Standalone headsets like Meta Quest 3 and PICO 4 ship the largest installed base for indie and mid-budget VR, so most new commercial projects target them first. PCVR through SteamVR reaches Valve Index, Quest via Link, and Windows Mixed Reality devices, with more headroom for fidelity. PSVR2 sells through one storefront on PS5, with strict certification and a single hardware spec. Multi-platform builds are common but add cost: a Quest-first build can usually be ported to PCVR with moderate effort, while a PSVR2 port often requires reworking input and certification material. The table below summarises the practical trade-offs teams weigh in pre-production.

VR game development engine and SDK stack

Unity and Unreal Engine cover almost every shipping VR project. Unity is the default for standalone Quest because the Universal Render Pipeline scales well to mobile GPUs and the XR Interaction Toolkit handles controllers, hand tracking, and teleport locomotion with little glue code. Unreal Engine fits PCVR and PSVR2 projects where lighting and material fidelity are the selling point; its OpenXR plugin supports the major runtimes, and Blueprints accelerate interaction prototyping. OpenXR is the recommended runtime layer on both engines because it isolates the project from vendor-specific code paths. Pick one engine, commit to OpenXR, and layer device-specific SDKs (Meta XR SDK for Quest-only features, PSVR2 SDK for eye tracking) only where the feature genuinely requires them.

The SDKs most VR teams use day to day:

Design for comfort: locomotion, vignette, frame pacing

Motion sickness in VR comes from a mismatch between visual motion and vestibular input. The fix is a combination of locomotion choice, on-screen mitigations, and stable frame timing. Teleport locomotion is the safest default for general audiences because it removes continuous visual flow. Continuous (smooth) locomotion suits action and shooter genres, but it should ship with a tunnel vignette, snap turn, and an optional comfort mode by default. Room-scale play stays comfortable because the player moves physically. Frame pacing matters as much as locomotion: a dropped frame on a 90 Hz headset is felt immediately. Reprojection (Asynchronous Space Warp, motion smoothing) is a safety net, not a target. Comfort is a shippable feature and should be tested with users outside the core team.

Locomotion patterns compared:

VR game development performance budget on standalone

Standalone VR is a mobile platform with a head-mounted display, and the budget reflects that. Quest 3 targets 72, 90, or 120 Hz; most shipping titles lock to 72 or 90 Hz and treat 120 Hz as a stretch goal for low-complexity scenes. The frame budget at 90 Hz is roughly 11 ms per eye, which the GPU and CPU share. Draw calls, overdraw, real-time shadows, and post-process effects are the usual cost drivers. Static batching, GPU instancing, baked lighting, single-pass instanced rendering, and Application Space Warp are the standard tools to hold frame time. Profile on the device early, not in the editor. PCVR has more headroom but the same principle applies: a stutter is a bug, not a frame-rate hint.

Reasonable starting budgets for a Quest 3 title (adjust per scene complexity):

Build input and interaction: controllers and hand tracking

Controllers are still the primary input on every shipping headset, but hand tracking is now reliable enough for menus, social experiences, and casual gameplay on Quest 3 and PICO 4. Design input as a layered system: the core verb (grab, throw, aim, push) should work on controllers first, then map to hand tracking where the genre allows. Haptics are part of the feedback loop, not a polish item. Sense controllers on PSVR2 add adaptive triggers, which can be used for tension and force feedback in archery, climbing, and weapon handling. UI is best in world-space with a ray-cast pointer for distant elements and direct touch for nearby panels. Accessibility settings (handedness, height, snap turn) belong in the first-launch flow.

Practical interaction patterns to support from day one:

Add mixed reality and passthrough where it fits

Mixed reality (MR) blends camera passthrough with virtual content on Quest 3, Quest 3S, and PSVR2. Meta's Scene API and Depth API allow virtual objects to react to walls, floors, and real furniture. MR is most useful for fitness, tabletop, education, and casual social genres where players want to stay aware of their real space. It is less useful for cinematic and immersive worlds where the player wants to be elsewhere. Plan MR as a mode, not as the whole game, unless the design genuinely needs the real room. Performance budgets for MR are tighter because passthrough composition costs frame time, so reduce scene complexity when MR is active.

Submit to the store: Meta Horizon, Steam, PSN

Each storefront has its own submission flow, comfort rating, and certification depth. The Meta Horizon Store now covers what used to be the main Quest Store and App Lab; submissions are reviewed against Meta's VR App Review (VRC) checklist for performance, comfort, and content. Steam has a lighter approval process but still expects clear VR support metadata and accurate input mapping. PlayStation Store certification for PSVR2 is the strictest, with mandatory checks on TRC compliance, save data, trophies, and accessibility. Plan a submission buffer of four to eight weeks for Meta and PlayStation, less for Steam. Translate and localise store assets early; VR storefronts surface trailers and immersive previews more than flat screenshots.

Store submission checklist used in practice:

Team and timeline for a commercial VR title

A small commercial VR title typically ships with a core team of eight to fifteen people: a producer, a technical lead, two to three engineers, two to four artists, an animator, a sound designer, and a QA engineer with a VR device matrix. Larger projects scale the engineering and art teams in parallel. A first prototype (vertical slice) usually takes three to four months on Quest-first scope. Production for a six to ten hour single-player VR game runs nine to fifteen months from greenlight to certification, with another four to eight weeks for store review and launch. Social and live-service VR titles need a smaller launch scope and a longer post-launch roadmap. Plan the QA budget around a device matrix that covers every supported headset, including the lead device and every secondary platform.

About Game-Ace

Game-Ace is a custom game development company founded in 2005, with 120+ in-house specialists and 200+ delivered games. The team works under Team Extension, Co-development, and Full-cycle delivery models with strict NDA and IP transfer. For VR scope, the studio provides virtual reality game development across Unity and Unreal Engine with OpenXR, supports Meta Quest, PSVR2, and PCVR targets, and handles comfort design, performance profiling, hand tracking, mixed reality, and store submission. Selected VR and immersive case studies are available in the Game-Ace portfolio; further references are shared under NDA. To scope a VR project or extend an existing team, contact our team with a short brief: target device, play model, expected session length, and the desired launch window.