GAMES · SOURCE CODE · VISUAL INSPIRATION
Scroll through actual gameplay and running-app images. Every available view is shown here, with its source and license beside the project.
This is a curated search snapshot, not an exhaustive index of every repository. Only reviewed gameplay or running-app imagery appears in this gallery. Projects without suitable images remain listed under All projects. No covers are used as gameplay. Publisher captures may depict different releases or devices.
CODE YOU CAN INSPECT
CHOOSE A VISUAL DIRECTION
Both reference screenshots are shown for every style. These are visual references, not open-source asset grants. Use the menu to return to reusable code.
FROM REFERENCE TO YOUR OWN PRODUCT
The main code license is the first check. Art, SDKs, models and paid services can carry separate terms.
Commercial derivatives—including closed-source games—are allowed under these licenses when their requirements are met. Keep the required license/copyright notices. Apache also has change-marking, NOTICE and patent provisions; BSD prohibits implied endorsement.
Promising bases: First Hand, The World Beyond, Phanto, Cryptic Cabinet, Spirit Sling, MRUK, MR Motifs, Meta XR Tabletop Games and Open Brush.
These can support a commercial Unity/Quest product within their permitted uses. They are not unrestricted asset-pack licenses. Check SDK sample-code scope, supported products, required notices, experimental features and third-party terms. Unity Companion content requires valid Unity licensing and its Unity-use conditions.
GPL permits selling software, but recipients retain GPL rights and corresponding-source requirements apply. ARC Car Racing is GPL 3.0. Review the combined Unity/Meta/Photon dependency situation before selecting it for distribution; a GPL label alone does not settle that compatibility.
Lasertag’s current license is PolyForm Noncommercial; older search results still say MIT. Commercial reuse needs a separate grant. BinaryPathway says MIT in its README but lacked a standalone license in this check; Oculus Passthrough Demos lacked a clear license. Resolve the grant before relying on either.
An independently implemented mechanic can be a starting point. Rewriting the code does not grant rights to reuse someone else’s artwork, music, characters, dialogue, level content, branding or other protected expression. Build your own identity and use your own or properly licensed assets. Other rights, including trademarks and patents, can also matter.
The US Copyright Office distinguishes game ideas and methods from protected artistic or literary content. That distinction is useful; it is not clearance for a particular clone in every jurisdiction. Read the primary guidance ↗
Kenney — CC0 asset-page downloads, especially simple props and environments. Quaternius — check each CC0 pack and its free/paid editions. Poly Haven — CC0 realistic models and materials; optimize for standalone Quest.
Research checked 26 September 2026 against repository READMEs and root license files. This is license identification, not a complete file/dependency audit. Preserve the exact licenses from the release or commit you ship. Workspace requirements for future implementation remain in the local Quest policies; none of these existing binaries is certified against them.
MECHANICS × VISUAL LANGUAGE
A genre describes what the player does; a visual style describes how it looks. The colored tags link the two. These are suggested combinations, not exclusive rules. Screenshots are existing gameplay references; commercial games are not automatically reusable source.
STANDALONE QUEST 3 · VR + MIXED REALITY
All style ratings here are engineering estimates for an optimized implementation. No project in this library has been benchmarked on a physical Quest 3 for this study. A screenshot cannot establish frame rate.
A small scanned environment or a detailed mechanical prop can look convincing through texture quality, correct scale, baked indirect light, normal maps and well-chosen reflection captures. Static geometry and limited viewpoints make this easier. Large unrestricted worlds with crowds, dense foliage, dynamic global illumination and many moving shadow-casting lights are a different workload.
At close range, silhouette quality, texture density and stable highlights matter. Baked lighting does not update like fully dynamic light: moving an object or changing the room can expose the illusion. Passthrough also needs believable placement, occlusion and contact cues.
Practical starting point Opaque flat, low-poly, painted, toy, miniature or neon-core scenes can leave a useful budget for gameplay.
Needs careful budgets Scans, plausible PBR, transparent crystals, ink lines and animated effects can work when their cost is controlled.
Substantial redesign from reference PC cinematic density, rich volumetrics, extensive destruction or complex organic surfaces require mobile substitutes. The aesthetic is still possible; the pictured implementation is not the proposed Quest build.
Opaque VR draws the environment. MR often draws fewer virtual objects, but passthrough composition, real-world occlusion and scene features still have costs. Measure the actual mode. With Link, Air Link or streaming, a PC renders the scene; that does not establish standalone headset performance.
Application SpaceWarp can turn a lower application rate into a higher display rate using synthesized frames. A 36 FPS application presented at 72 Hz is not 72 newly rendered application frames. North Star and Unreal Showdown should be checked for their selected SpaceWarp settings before using them as performance references.
Build Android for the headset, choose 72 Hz explicitly, and test the busiest gameplay with the intended render resolution and anti-aliasing. Inspect CPU and GPU timings separately; their work overlaps, so do not simply add them. Check dropped frames, thermal behavior after a sustained session, memory, loading hitches and both eyes. Leave headroom below the 13.89 ms deadline.
Start with baked lighting, culling, LOD, pooled objects, shared materials and limited transparent effects. Use foveation and resolution scaling where appropriate, then inspect their visible quality. There is no universal polygon count that guarantees 72 FPS.
QUEST MARKET SNAPSHOT · 26 SEPTEMBER 2026
Observed first 24 entries in Meta’s Top-earning this week collection. Prices are the EUR amounts displayed in this browsing session on the English (UK) store. This is a dated, regional sample, not a census or a live price feed.
The nine paid downloads in this sample span €14.99–€39.99 at the observed checkout offer, with a €17.99 median. Their displayed regular prices have a €24.99 median; four are discounted. A roughly €15–€30 paid base price is a useful initial comparison range for this particular group, not a proven optimal launch price. Premium releases elsewhere in the store, subscriptions and add-ons are outside these statistics.
Free download does not mean no monetization. Revenue rank can include spending after installation, but this collection does not reveal the mix. Public rank and review counts do not provide copies sold or revenue totals. Pricing alone cannot predict profitability.
| Revenue rank | Game / genre | Observed EUR price | Visual aesthetics |
|---|
Publisher screenshots from the linked Quest listings. Visual tags are editorial descriptions, not store categories or engine identifications. Screenshots not obtained are explicitly marked in the table; no covers are substituted.
ORIGINAL PROPOSALS · UNBUILT
Sixteen design proposals to explore beyond the existing catalog. They combine mechanisms and visual treatments in potentially useful ways; novelty has not been established by an exhaustive prior-art search.
UNITY VS UNREAL · FINAL HEADSET IMAGE
Unreal Engine also supports standalone Quest 3 VR and passthrough. Neither engine has a mandatory visual style or an automatic realism advantage on this headset. With comparable art, lighting, shaders, resolution and rendering cost, both can produce similar kinds of final images. This is a technical assessment, not a measured same-scene A/B benchmark.
| What you see | Unity on standalone Quest 3 | Unreal on standalone Quest 3 | Practical consequence |
|---|---|---|---|
| Overall visual character | URP supports realistic materials and custom stylized shading. | Mobile materials support realistic and stylized shading too. | Art direction, texture choices and lighting dominate the “look”; engine branding is not an image-quality setting. |
| Lighting and shadows | Baked lightmaps and probes with selective real-time lights are a useful baseline. | Precomputed lighting and reflection captures suit Mobile Forward; select dynamic lights carefully. | Both can show convincing static bounce light. Moving light, dynamic shadows and changing environments consume more budget. |
| Surface detail and reflections | PBR textures, normals and reflection probes; simplify layers for URP mobile shaders. | PBR mobile materials and reflection captures; material complexity must fit mobile limits. | Close objects can look rich in either engine. Texture quality does not replace silhouette geometry; baked reflections do not behave like ray tracing. |
| Edges and clarity | Unity recommends MSAA and single-pass/multiview for untethered XR. | Mobile Forward supports MSAA; Meta’s common Quest configuration uses HDR disabled. | Resolution, MSAA and foveation affect clarity more directly than choosing one engine. Test shimmer on thin lines and highlights. |
| Glow, filmic effects and grading | Unity 6.3+ offers selected on-tile post-processing effects; older versions have different cost guidance. | Epic’s head-mounted Mobile Forward/HDR-off feature table excludes many desktop post-process effects. | Desktop previews can change noticeably on the headset. Budget effects deliberately and check the exact engine/fork version. |
| Nanite, Lumen, ray/path tracing | Do not use a desktop HDRP/path-traced picture as a standalone URP target. | Epic’s mobile renderer matrix lists Nanite, Lumen and hardware ray tracing as unsupported. | Switching to Unreal does not bring those desktop features to a shipping native 72 FPS Quest scene. |
| Camera passthrough | Meta XR integration supplies the device’s passthrough composition and MR features. | Meta XR plugin supplies passthrough composition and MR features. | The engine does not upgrade the physical cameras. Object integration, lighting approximation and depth masking affect believability. |
| PC VR / streaming | PC render paths can use a different, more expensive feature set. | Desktop rendering can support features absent from the mobile renderer, subject to VR support and cost. | Rendering occurs on the PC. Image compression, streaming and latency become additional factors; this is a separate product target. |
Unity is the practical first option for this catalog because the 61 audited source projects here are Unity-oriented, including complete hand-tracking and MR game samples. That recommendation follows the available starting code, not a claim that Unity renders faster in every scene.
Unreal is a sound option if your team already works in C++/Blueprints or has Unreal art and tooling. Evaluate Meta’s engine fork and plugins against your required version. Importing a desktop Unreal scene still requires deliberate mobile lighting, mesh and material work.
Use the same legally reusable room, identical camera pose, mesh LODs, textures and lighting intent. Match eye-buffer resolution, foveation, refresh rate and MSAA. Compare an optimized Android build from each engine on the same warmed-up Quest 3, recording CPU/GPU time and headset captures. Then compare the best image each can sustain within the same 72 FPS budget.
We have not run that controlled test. Unrelated game screenshots would confound engine, budget, artist effort and capture platform, so they cannot establish an engine winner.
These are additional official source-available samples. They are outside the 61-project Unity license audit. Meta SDK and Epic license terms apply; a public GitHub repository does not by itself make its content permissively licensed.
Background composition, styling, color LUTs and masking. Useful for learning the MR rendering path.
GitHub source ↗ · LicenseScene-aware placement and utilities on top of Scene API.
GitHub source ↗ · LicenseA PC VR scene ported to Quest, with profiling controls including SpaceWarp, foveation and MSAA. It is an optimization case study, not proof of native 72 FPS at every setting.
GitHub source ↗Engine licensing is separate from sample-code licensing. Epic’s current EULA includes revenue-based royalties, exclusions and reporting requirements, including a specific Oculus Store/successor revenue exclusion. Use the actual agreement for a commercial release; do not infer engine rights from an MIT sample.
BUILD YOUR OWN VISUAL IDENTITY
These supply reusable content for your Unity project. Each pack still needs suitable materials, colliders and game logic.
Simple 3D props, environments, vehicles, UI and audio. Asset-page downloads are CC0.
Browse asset packs ↗ · License informationCharacters, modular worlds, props and animation libraries. Check each pack’s CC0 grant and the content included in its free or paid edition.
Browse asset packs ↗CC0 models, PBR materials and HDRIs for realistic surfaces. Optimize meshes and textures for standalone Quest.
Browse assets ↗ · License informationQUEST 3 DESIGN LAB · LARGE IMAGE EDITION
27:11 · 7 chapters · 64 comparison-led sections. Choose natural stock Kokoro male or female narration. Both editions show the same complete collection, with larger images, distinct games in each comparison, and a four-second chapter break before AI proposals.
The opening credits the AI workflow, Hyperframes and Kokoro. All 16 proposed concepts now have original AI-generated snapshots. These are unbuilt design visualizations; existing-game captures remain separately credited.
Stock voices, not celebrity imitations. Select an edition, then press Play.
Code permissions do not automatically cover artwork, branding, SDKs or other dependencies. Copyleft can permit commercial sale with source-sharing obligations.
Loading topic index…
61 code projects, including 48 with open-source main code and 47 permissive licenses; 34 styles; 26 genres; 24 market examples; 16 original concepts.
All 214 collected image references appear, including 33 animated running-app references, alongside the new concept art. Related examples share large comparison frames. Each frame shows a game only once; additional shots advance on later pages.
AI concepts show a proposed visual direction. PC-store captures are labeled as references. Performance ratings are design estimates, not physical Quest benchmarks.
Research and EUR prices: 26 September 2026. Caption chunks use approximate phrase timing. Original instrumental music accompanies both voices.
This is an AI-generated video, made with GPT-six Astra in Extra High mode, Hyperframes and Kokoro. Here's what that workflow can create. Let's explore the possibilities for Quest.
Imagine pulling a ribbon of wind through your living room, or finding a forgotten city by listening to a stone. That's the interesting part of building for Quest: deciding what a person gets to do. We're going to compare the looks, the games, and the code that could make those moments possible. Then we'll explore sixteen original concepts, with newly generated pictures that show the idea in action.
Start with where the experience lives. Full virtual reality lets you control the whole world: its lighting, scale and atmosphere. Mixed reality keeps the real room and adds things that seem to belong there. A tiny strategy board can work beautifully in either. A dark horror corridor is much easier when you control the surroundings. Showing passthrough also doesn't automatically give your app access to camera images; that's a separate capability.
A quick key before we begin. Green brackets mean permissive source code. Blue linked arrows mean open code with share-alike obligations. An amber tag means platform-specific terms, and red N C means noncommercial. A question mark means the license needs clarification. Those badges describe the main code. Game screenshots are visual references, and the purple diamond marks an invented concept. I'll leave the detailed license names on screen so we can spend our time on the design.
Look at the difference between a convincing object and a dense world. The Room puts material detail where you can inspect it. Moss uses designed shapes and lighting to make its world feel believable. Alyx adds a much larger density of detail, but these PC images aren't a standalone Quest benchmark. For your own project, choose the few surfaces people will touch, and spend detail there. Realism becomes manageable when the scene has a clear boundary.
Firewatch, The Last Campfire and Dordogne all guide your eye with color, but they do it differently. One builds broad landscape shapes. Another feels like a painted storybook. Watercolor introduces soft pigment edges. You can borrow the principle without borrowing anyone's artwork: decide which color means safety, curiosity or danger. On Quest, authored textures and lighting can carry that mood while keeping expensive full-screen effects under control.
A miniature board invites you to lean in. Clay invites touch. Paper makes folds and silhouettes part of the character. Toy plastic makes a world feel assembled and playable. These references span Demeo, Hylics, Paper Beast and Bricktales; they aren't one interchangeable style. Pick a material language, then make its scale consistent. A paper bridge should bend like paper, while a chunky toy bridge can snap together with a satisfying click.
These four directions make different uses of less detail. Faceted geometry exposes the shape. Minimal geometry removes distractions. Flat action art separates a threat from its background. Graphic noir uses a tightly controlled palette to steer movement. Compare Grow Home, Cubism, Superhot and Pistol Whip: the useful question is what becomes easier to read. A simple scene can look deliberate and rich when its silhouettes and contrasts have a job.
A cartoon, a graphic novel and an ink painting can all exist in three dimensions. Hi-Fi Rush favors clean shaded forms; Sable emphasizes line and open color; Okami brings brush-like marks into the picture. For a headset, those marks must survive stereo viewing and movement. Very thin outlines can shimmer. Make the important contours bold, test them at real viewing distance, and let a few authored marks do more than a screen full of effects.
Retro isn't a single filter. Block-built worlds, low-resolution models, pixel textures with modern lighting, and gritty early-three-D spaces each remove different information. Teardown, A Short Hike, Valheim and Dusk show those differences. Don't mistake chunky art for cheap simulation: thousands of breaking blocks can overwhelm a processor. Decide whether the nostalgia comes from shape, texture, lighting or motion, and keep the parts that help the game read.
Beat Saber makes luminous targets readable. Rez uses a translucent, data-like world. Tetris Effect builds a sensory atmosphere around a clear play area. That's three different jobs for light. In your headset scene, protect the important action from the spectacle. A few opaque emissive shapes can suggest energy beautifully; stacking transparent particles across both eyes is where the cost climbs. Let the music drive a small number of intentional changes.
Obra Dinn's monochrome patterning, Manifold Garden's impossible architecture, and Squingle's glass-like forms create very different kinds of uncertainty. The pleasure comes when you learn their visual rules. In VR, preserve depth cues even when the world is surreal. For a liquid-crystal look, one small reflective-looking object is a much friendlier starting point than layers of full-screen refraction. Mystery works best when the next useful action remains visible.
Rounded pastel forms make Slime Rancher inviting. Gothic shapes in Lost in Random feel curious and uneasy. Organic surfaces in Scorn can make familiar anatomy seem alien. Compare the proportions, edges and material response, not just the brightness. You can create tension with silhouettes, pacing and sound long before adding dense surface detail. These are art references from their stated platforms, not promises that a desktop scene will fit unchanged on Quest.
Scanned miniatures and scenic realism start with the same attraction: a place that feels specific. Puzzling Places turns that detail into an object you can assemble. Brink Traveler presents landscapes you can inhabit. The difference in scale changes the budget. A bounded miniature can keep detail near your hands; a landscape must manage distant geometry and textures. Clean up scans and choose viewpoints deliberately instead of importing raw capture and hoping.
Sprites in three-dimensional space and hard-surface cockpit graphics solve different problems. Octopath's layered picture gives depth to deliberately simplified characters. Battlezone uses bold edges and cockpit-like forms to communicate direction and machinery. In VR, a flat sprite needs careful orientation, and a cockpit label must remain readable at its actual distance. These techniques become useful when they clarify depth or operation, rather than simply decorating a familiar genre.
Shooters, melee, rhythm and fitness all ask you to act at the right moment. Compare their targets and reach zones before their scenery. A neon rhythm cue, a flat-shaded opponent and a grounded weapon can all work, provided the player sees the important motion. Give people seated or limited-reach options where the activity allows it, and measure tracking, physics and audio timing as part of the experience.
Spatial puzzles, escape rooms, reconstruction and stealth share a useful habit: they reward looking carefully. Cubism and Squingle make shape the puzzle; a mechanical escape room makes the object itself a clue. Stealth shifts attention to routes and timing. These are strong candidates for contained environments, where a few readable props can carry a lot of play. Avoid clues that only work when the player presses their face against them.
Platforming, narrative adventure and exploration differ in what movement is for. A miniature character crossing a gap can be exciting while the viewer stays seated. A story can pull you through a sequence of places. Tourism lets the place itself be the reward. Choose locomotion after deciding that purpose. Full VR can carry you into a landscape; mixed reality can put a whole journey on the table.
Tabletop tactics, strategy, construction and management make systems tangible. You move a unit, connect a machine or reshape a small world, then watch the consequence. Miniatures, paper and toy materials are useful here because scale becomes legible. The quiet danger is simulation cost: a simple-looking board can still contain hundreds of thinking agents. Bound the active pieces and make each change easy to understand.
Survival, horror, creature care and repeatable runs can all create a reason to return, but they create very different feelings. Survival makes resources matter. Horror makes uncertainty matter. Creature care makes attachment matter. A roguelike makes each new attempt a different decision. Choose that emotional loop first, then mix the visual language around it. A rounded little world can still have meaningful stakes without becoming visually aggressive.
Creative tools, social spaces, training and calm sensory experiences belong in this survey too. Open Brush shows how movement can become creation. A learning app might need one accurate mechanism more than a spectacular room. A social activity needs shared attention and understandable gestures. A quiet experience needs pacing that gives people time to notice. The design question is what the person accomplishes, feels or understands after using it.
Sports, racing and flight place unusual weight on consistency. A beautiful course loses its value if the ball or steering response feels wrong. Cockpits need readable instruments; races need stable distant scenery. The driving and flight references shown here include PC examples, so use them to study composition and control, not to estimate a standalone frame rate. A tabletop racer is another route when you want mixed reality without full-speed virtual travel.
Now let's check the market without pretending a ranking tells us everything. Our dated sample contains twenty-four ranked Quest apps: fifteen free to enter, and nine paid. The displayed paid prices run from fourteen ninety-nine to thirty-nine ninety-nine euros. Their median displayed price is seventeen ninety-nine; the median list price is twenty-four ninety-nine. Offers change, and this is not the average of the whole store. Free entry also doesn't mean the developer earns nothing.
These four ranked examples put movement and social energy ahead of intricate materials. Hero X Hero, Scary Baboon, U G and Gorilla Tag have different identities, but their accessible silhouettes and expressive movement are worth comparing. The lesson isn't to copy one successful character. It is to make the core action understandable quickly, then give friends a reason to keep playing together.
VRChat, Vegas Infinite, Yeeps and Orion Drift organize social play around very different spaces and activities. Their value can come from who is there and what people make happen together. Visual consistency still matters, especially when users create content or many avatars share a room. The free-entry label here describes the displayed storefront price, not a conclusion about their entire business model.
Beat Saber, Golf Plus, V R F S, Gym Class and N F L Pro Era show how a recognizable physical activity can anchor an app. Compare the clarity of a target, a ball and a playing area. Some are paid, some free to enter. Price alone doesn't explain that difference; content, social features, licensed material and ongoing services may all affect the offer.
Blade and Sorcery, Bonelab, Beyond Sandbox, Forefront and Ghosts of Tabor all sit near action or simulation, but they offer different kinds of play. A physics sandbox and a tactical match don't need identical art direction. Their displayed prices span much of our paid sample. Use this comparison to define the experience you're selling; don't treat the highest price as a target your project automatically deserves.
Animal Company, Downshot, Animal Rivals, Zombonk, I Am Cat and Lethal Ape fill out the sample. They show how strongly a clear role or activity can communicate an experience. The playful, scary and action-heavy looks aren't confined to one business model. These examples are a market snapshot, not endorsements or a prediction of demand. The useful output is a shortlist of audiences and experiences to investigate further.
Tap the stone, then listen as a buried room starts to appear. Acoustic Archaeology turns sound into a way of exploring space. The chalk contours make discovery visible without revealing everything at once. I'd prototype one small excavation with three kinds of response, so the satisfying moment comes from learning what each sound means.
Here, wind is something you can braid. In Weather Weaving, your hands pull textile ribbons around a little island, steering boats and seeds toward their destinations. The woven-cloud look makes an invisible force tangible. Start with a few authored flow paths; the puzzle should be about shaping the wind, not fighting an unpredictable simulation.
This music box remembers more than one version of the past. Memory Strata lets you peel away paper, paint and wood to discover contradictory scenes. Each material becomes a different kind of memory. Keep the first story small: one object, three layers, and a reveal that makes you reconsider what you saw before.
Instead of planting something, make room for it. Negative-space Gardening grows life inside the spaces you carve from ceramic shells. The contrast between a quiet solid surface and a living interior is the whole attraction. A bounded tabletop volume would let you explore that idea without promising an infinitely editable world.
Draw a route, and the map rises under your finger. Tactile Cartography turns paper contours into paths for tiny travelers. A river becomes a real obstacle; a ridge becomes a decision. The brass edges keep the geography readable. Begin with a fixed terrain grid and a handful of travelers, then see which gestures feel most satisfying.
Try solving a problem together without speaking. Social Shape Language gives two players a small vocabulary of soft objects they can stretch and connect. A loop might invite someone to join; a bridge might offer help. The interesting work is making those meanings discoverable. Six shapes and a few shared tasks would tell you much more than a huge world.
This little habitat tells you how it feels through its glaze. Living Enamel turns care into changes in color, cracks and pattern, so the surface becomes part of the creature's expression. It could feel precious without needing realistic fur or complicated glass. One creature, one habitat and a few clearly readable moods make a useful first test.
Pull a bright thread, and a bridge starts to hold its shape. Woven-light Architecture makes tension visible through thick, interlaced strands. A sagging section asks for adjustment; a balanced junction becomes a little reward. Keep the strands bold enough to survive headset resolution, and make the structure understandable before filling the room with it.
A scattered set of leaves becomes a flower when you find the right view. Viewpoint Herbarium uses physical paper-like layers to turn observation into a puzzle. The key is that the alignment must make sense to both eyes. Three carefully authored plants would be a better test than a garden of optical tricks that only works in a flat screenshot.
Your coffee table is a meadow, the shelf is a cliff, and the sofa arm has become a pond. Room Ecologies connects a few small habitats across a real room. You change a rule and watch the creatures migrate. Keep the real furniture visible, make the routes readable, and treat room mapping and persistent placement as separate things to validate.
Move the linkage just right, and it answers with music. Constraint Choreography turns a small mechanical puzzle into a duet between your hands. Color and softly lit joints show which relationships are working. The first version only needs a few parts and clear musical responses. Design its reach for seated play from the beginning.
These societies need different things because they're made from different materials. Cloth catches wind; stone gives weight; wood spans a gap. Material Diplomacy turns negotiation into a visible physical problem on one shared board. The appeal is finding a trade that changes everyone's possibilities. Three factions would give the idea enough tension to test.
Rotate the object, and its shadow becomes a path. Shadow Loom gives a tiny traveler a world built from the silhouettes you arrange. Warm porcelain and dark paper-like shapes keep the cause and effect clear. Begin with one controlled lamp and a few props. Real-room shadows are much less predictable, so don't base the first puzzle on them.
In Palette Economics, color is a resource you can spend. Give red to a bridge and it becomes usable, but another district has to give that color up. The changing picture is also the changing strategy. Each resource needs a shape as well as a color, so the rules remain understandable to players who distinguish colors differently.
One player can move the crane; the other can see the tiny landing platform. Scale Relay makes cooperation depend on different perspectives. The giant view and the inhabitant's view each reveal something useful. A single table and two fixed scales would let you test the shared task without immediately tackling free-form size changes and uncomfortable camera motion.
A clean crescent in the dust tells you an object has moved. A worn hinge suggests how it was used. Patina Detective turns material history into evidence you can handle. The realism has a purpose: helping you infer an action. Build one small mystery around a few readable marks, then test whether people notice the clues at normal headset distance.
Now we can choose code by the interaction we want. First Hand is useful for learning hand-driven actions. The World Beyond brings a character into a room. Phanto turns room geometry into action, while Cryptic Cabinet builds a puzzle experience around it. Compare these as complete examples to learn from, then extract the smallest interaction that serves your own idea.
Spirit Sling gives you a social tabletop reference. Discover explores a collection of shared mixed-reality activities. Spatial Lingo connects real objects with language practice. These are three different product directions, even when they share room-aware foundations. Before adapting one, identify which networking, account or external-service setup it expects; those dependencies can shape the size of your first prototype.
The utility-kit samples, M R Motifs, interaction examples and starter scenes are useful when a complete game is too much to inherit. Think of them as small answers to recurring problems: placing something, touching it, moving between real and virtual space. Choose the example that demonstrates the exact behavior you need, and match it to the package versions in your project.
Depth, camera access, shared anchors, room-data sharing and multiplayer alignment solve different problems. An object disappearing behind your sofa is occlusion. Recognizing something in a camera image is vision. Two people agreeing where a board sits is spatial alignment. Keep those responsibilities separate in your design, so a failure in one system doesn't become a mysterious failure of the whole experience.
A tracked keyboard, an avatar interaction sample, body movement and Gorilla Locomotion point to very different relationships with the body. You might be typing at a desk, expressing a gesture or moving through a world with your arms. Borrow the interaction model carefully. The locomotion repository is not a license to copy the complete Gorilla Tag game or its characters.
North Star, Move Fast, Ultimate Glove Ball, Decommissioned and Shared Spaces give you larger systems to inspect. Watch how movement, social presence, a goal and feedback work together. A showcase can teach a production technique without being the fastest starting template. A smaller multiplayer sample can teach a clean shared loop. Pick the lesson you need before taking on the whole project.
Spatial audio and haptics can make a modest-looking interaction feel much more convincing. The virtual-camera and voice examples address capture and spoken interaction instead. These tools solve different parts of the experience, so add them with a specific purpose. A well-timed response to touching an object is often more valuable than adding another visual effect.
Compositor layers, performance settings, SpaceWarp, asset streaming and shader warm-up each address a different rendering problem. They aren't interchangeable performance switches. Keep important interface content clear, avoid first-use stalls, and control what is loaded. Reprojected frames are also different from rendering every frame natively, so keep that distinction visible when you evaluate a seventy-two-frame target.
Open Brush is a substantial creative-tool foundation. The tabletop examples explore smaller mixed-reality game loops. Arc Car Racing and Tinker X R add different ways to play and construct. Compare the interactions and the corner badges together: a project can be open source and still have obligations that affect how you distribute your changes. The detailed table remains in the library.
Bracer, room scanning, camera wrappers and marker tracking can support very different applications, from spatial interfaces to recognizing a physical reference. The important distinction is what data each tool actually supplies. A camera image, a room mesh and a tracked marker are not the same coordinate system. Start with one visible, testable result before combining them.
These camera tools cover utilities, reconstruction, streaming and computer vision integration. A useful prototype might simply place one result correctly in the room. Be especially careful with examples that wrap a separate paid dependency; the badge describes the repository's main code, not every component it needs. Check the pipeline end to end on the device before building a large experience around it.
Synth V R and the D A W show how a spatial interface can organize sound. The mixed-reality template offers a smaller starting shell. Those are useful contrasts: inherit a full application's structure when its interaction is close to yours, or start small when you only need the basic environment. More existing code is helpful only when you understand what you are carrying forward.
Mixed Reality Toolkit, Ultimate X R, V R T K, Unity's interaction examples and A R Foundation cover overlapping territory with different assumptions. Compare supported devices, input patterns and package maintenance before choosing one. Don't stack several large interaction systems just because each has an attractive demo. A coherent setup is easier to debug and easier to bring to a dependable release.
MiniGolf and the older passthrough examples can still teach a focused loop or an integration pattern. Their age is part of the evaluation, though. Check the editor version, packages and platform APIs before treating an older repository as a current starter. Sometimes rebuilding the small behavior in your current framework is simpler than carrying forward the entire old project.
Lasertag carries a noncommercial restriction in this audit. BinaryPathway and the older passthrough demos have unclear licensing in the collected record. They can be interesting to study, but don't assume you have permission to turn their code into a paid product. This is why the badges stay with each project: they keep an exciting visual from quietly turning into a reuse assumption.
At seventy-two frames per second, a frame arrives about every thirteen point nine milliseconds. That budget includes more than the art. Your processor handles scripts, physics and draw preparation; the graphics chip handles resolution, materials, lighting and overdraw. Measure both on the headset after it warms up, in the busiest part of the scene. A pretty empty-room test won't tell you what happens when players and effects arrive.
Quest can produce very convincing materials and carefully staged places. The trick is choosing what earns that detail. Bake lighting where the scene allows it, use reflection captures or authored highlights, simplify distant geometry, and keep transparent layers under control. A beautiful inspectable object is a different problem from an enormous dynamic city. Our practical and conditional labels are design estimates; only your actual headset build can tell you whether the target is met.
Neither engine owns realism, cartoons or a particular color palette. Similar content and lighting can lead to similar pictures. The difference is the rendering path, supported features, tools and the work your team must do. Unity's mobile render pipeline and X R ecosystem offer many examples in this library. Unreal also needs its supported mobile path. A desktop cinematic using expensive features is not what automatically appears when you package for Quest.
A fair engine comparison uses the same scene, headset, resolution, lighting intention and performance target. Compare the final packaged builds, including frame timing and the features you had to replace. We haven't run that matched benchmark here, so there is no honest universal winner to announce. Choose the tools your team can use to make the experience clear, stable and maintainable.
Choose one activity, one visual family and one code foundation. Build the smallest moment that makes you want to do it again: a satisfying grab, a clever shadow, a creature responding to care. Then test it in the headset and with another person. The library keeps the full screenshots, repositories, prices and license details together, so you can return to any comparison and turn the idea that caught your eye into a concrete next step.
20 SCIENCE-FICTION DIRECTIONS · ORIGINAL CONCEPT STUDY
Each world starts with an interaction, a physical constraint and a visual direction. These are overlapping themes and subgenres, ranging from grounded engineering to explicitly speculative social premises.
AI-generated visualizations of unbuilt proposals, not captured gameplay or evidence of 72 FPS. The 72 FPS goal applies to a bounded prototype that still needs device measurement. Photorealistic room backgrounds illustrate passthrough; they are not virtual scene geometry.
Stock Kokoro male narration · original instrumental music · AI-generated concepts · Download captions
LOCAL APP OPPORTUNITIES · RESEARCHED 26 SEPTEMBER 2026
Proposals grounded in local services, places and reusable data. Each separates the verified Helsinki detail from the proposed experience, first prototype and possible business model. These are unbuilt ideas, not claims of market demand or global novelty.
MR means virtual objects in the visible room; VR means an authored virtual environment. Several everyday planning tools belong on a phone first. All Quest frame-rate statements below are design estimates for a bounded prototype, not measured 72 FPS results.
Small and Medium describe relative implementation scope, not delivery promises. Suggested buyers and business models are hypotheses to test; none of the named organizations has agreed to a partnership.
Helsinki 3D offers mesh and urban models under CC BY 4.0. Convert selected areas into headset-ready assets; downloads are not optimized game scenes.
HSL data generally uses CC BY 4.0, while OSM-derived data has ODbL conditions. Check API access requirements before implementation.
Service Map terms allow commercial data use with attribution, but explicitly restrict certain public-art photographs. Data and image rights differ.
HSY Waste Guide requires API registration and uses CC BY 4.0. FMI's open data has attribution and request-limit requirements.
Choose one audience and one recurring task. Show a small interactive prototype to a few intended users, observe where it helps, then verify willingness to pay or a partner's procurement route.
For Quest, build one complete interaction and measure it on a warmed-up headset. For mobile, test data freshness and whether the app saves steps compared with existing city services.