How to Make an App Like Among Us

A practical breakdown of the architecture, tech stack, and cost drivers behind building a real-time multiplayer social deduction game — from lobby systems to anti-cheat logic.

What Makes Among Us Technically Hard to Replicate?

Among Us looks deceptively simple. The art style is minimal, the maps are small, and the mechanics are straightforward. What makes it hard to build is the real-time multiplayer layer underneath, and the social deduction logic that sits on top of it.

The game needs to synchronise player positions at roughly 20–30 updates per second across up to 15 clients simultaneously. It needs a voting system that is authoritative, tamper-resistant, and resolves in near real-time. It needs proximity-based voice or text chat. And all of this has to work across iOS, Android, PC, and console, with players on wildly different network conditions.

Each of those requirements is a separate engineering problem. They compound.

What Tech Stack Should You Actually Use?

The choice of stack here has real consequences. Let us go through each layer.

Game Engine

Unity is the practical choice for this category of game. Among Us itself is built on Unity. It handles 2D sprite rendering, physics, and cross-platform export (iOS, Android, Windows, macOS, consoles) from a single codebase. Unreal Engine is overkill — it is built for high-fidelity 3D and brings a heavier runtime than you need. Godot is viable if your team already knows it, but has a narrower ecosystem for multiplayer tooling.

Unity's Netcode for GameObjects (NGO), introduced in 2021, is the current first-party option for multiplayer state synchronisation. For more control, Mirror (a community fork of the deprecated UNET) is battle-tested and has a larger set of production case studies for mid-scale multiplayer games.

Networking Model

Among Us uses a server-authoritative model for critical game state (tasks completed, kills, votes) and client-side prediction for movement. This is the correct split. Client-authoritative movement would make the game trivially exploitable; pure server-authoritative movement introduces perceptible lag.

For transport, you have two realistic options:

Option	Latency	Reliability	Use case
UDP (custom or via Netcode)	Low	Unreliable by default	Player movement, position
TCP / WebSocket	Higher	Reliable	Votes, task completion, chat
Relay + WebRTC	Variable	Configurable	Proximity voice

Photon Fusion (version 2, released 2023) is worth evaluating. It handles rollback netcode and interest management out of the box. At scale, its hosted relay infrastructure saves significant DevOps time, though the pricing model (based on peak concurrent users) needs to be modelled against your expected player counts before you commit.

Backend Infrastructure

You need a matchmaking service, a lobby service, a game session service, and a persistence layer. These are distinct responsibilities and should be deployed separately.

Unity Gaming Services (UGS) bundles Lobby, Relay, and Matchmaker as managed services. This is the fastest path to a working prototype. The trade-off is vendor lock-in and limited control over matchmaking logic. If you anticipate custom matchmaking rules (skill-based, region-weighted, or friend-group-aware), you will outgrow UGS matchmaking quickly and need to replace it with a custom service, likely written in Go or Node.js with Redis for session state.

For a game at Among Us's scale (peaked at around 500 million downloads as of 2023), the backend needs horizontal scaling and regional deployment. AWS GameLift, Agones on Kubernetes, or Edgegap are the serious options for dedicated server hosting. Edgegap is worth a look specifically because it deploys game servers on edge nodes, reducing latency for players in regions that cloud providers underserve.

Proximity Voice

This is frequently underestimated. Among Us's proximity chat mod (popularised by the community before being adopted officially) requires players to hear only those within a configurable radius, with audio attenuation. Native integration requires WebRTC-based spatial audio. Vivox (now Unity-owned) is the production-grade option. Agora and Daily are alternatives with more flexible pricing at lower user volumes.

How Do You Handle Game State and Anti-Cheat?

State management is where most indie multiplayer games break under moderate player load or adversarial input.

The game server must be the single source of truth. Every player action, from completing a task to calling a meeting, should be sent as an event to the server, validated, and then broadcast back to all clients. Clients render what the server confirms, not what the local player submitted.

Anti-cheat in this genre is a specific problem. Speed hacks, kill-anywhere exploits, and voting manipulation are the common attack vectors. Server-side validation catches most of these: if a kill event arrives from a player whose last known server position is more than a plausible distance from the victim, reject it. For task completion, the server should track task state independently and not trust the client's self-reported completion.

Rate-limiting events per player per second is a cheap and effective layer of protection. Temporal anomaly detection (flagging clients that report game-state changes out of sequence) catches a second class of exploits without requiring kernel-level anti-cheat software, which is not viable on mobile anyway.

Cross-Platform and Store Considerations

Building cross-platform with Unity gives you a single codebase, but store submission and platform certification are separate workstreams for each target.

Apple App Store requires compliance with App Store Review Guidelines 4.7 (game controllers) and 1.4.3 (account systems). If you use Apple Sign-In as one auth method, you must offer it wherever you offer other social logins. Google Play has its own Play Integrity API requirements for tamper detection on Android. If you are targeting consoles (Nintendo Switch was a major platform for Among Us), each platform has its own SDK, certification process, and multiplayer policy requirements.

Plan for approximately 4–8 weeks per console platform for certification, assuming no critical findings. Mobile review cycles are shorter but iterative rejections are common if the game has any user-generated content features.

What Does It Actually Cost to Build?

Rough order-of-magnitude estimates, based on a team with mid-to-senior experience:

Phase	Scope	Approximate Duration
Core game loop (local)	Mechanics, map, tasks	8–12 weeks
Multiplayer integration	Netcode, lobby, matchmaking	10–16 weeks
Voice chat	Proximity audio integration	3–5 weeks
Anti-cheat + validation	Server-side logic	4–6 weeks
Cross-platform build + QA	iOS, Android, PC	8–12 weeks
Backend infra + DevOps	Scaling, monitoring	6–10 weeks

A full production build targeting iOS, Android, and PC, with a team of six engineers and one producer, runs to roughly 12–18 months. That does not include live operations, content updates, or moderation tooling if you add user-generated content.

The fastest path to validating the core loop is to build mobile-first with UGS for matchmaking, skip proximity voice in v1, and defer console targets. This gets you to a testable multiplayer build in roughly five to six months.

Conclusion

The core game logic in Among Us is not what makes it expensive to build. The multiplayer infrastructure, the anti-cheat layer, the cross-platform certification work, and the operational tooling to run live sessions at scale are where the time and cost accumulate.

If you are evaluating this seriously, the next concrete step is to build a network architecture diagram that covers your session server, relay layer, matchmaking service, and persistence layer before writing a line of game code. Getting that topology wrong early is the most expensive mistake you can make in this category of project.

---

FAQ

How long does it take to build a multiplayer game like Among Us? A production-ready cross-platform build targeting iOS, Android, and PC typically takes 12–18 months with a team of six to seven people. A focused mobile-only prototype with basic multiplayer can be functional in five to six months if scope is deliberately constrained and you use managed services for matchmaking and relay.

What engine should I use to build a game like Among Us? Unity is the most practical choice. Among Us itself runs on Unity. It supports 2D development, cross-platform export, and has the broadest ecosystem of multiplayer tooling, including Netcode for GameObjects and Photon Fusion. Unreal Engine is disproportionate for this type of game; Godot is viable but has less mature multiplayer infrastructure.

How do you prevent cheating in a multiplayer social deduction game? Use a server-authoritative architecture where all critical game events (kills, votes, task completions) are validated server-side before being applied to game state. Rate-limit player events, validate positional plausibility for interaction events, and track task state independently on the server. Kernel-level anti-cheat is not viable on mobile platforms.

What does it cost to run the backend for a game like Among Us at scale? Costs depend heavily on concurrent player counts and regional distribution. At moderate scale (tens of thousands of concurrent players), managed services like Unity Gaming Services or Photon Fusion can cost a few thousand dollars per month. At high scale (hundreds of thousands of concurrent players), dedicated server hosting via AWS GameLift or Agones on Kubernetes is more cost-efficient but requires significant DevOps investment.

Do I need proximity voice chat in v1? No. Proximity voice is a meaningful feature but it is not part of the core game loop. Among Us shipped and became massively popular with text chat only. Adding proximity voice requires WebRTC integration, spatial audio handling, and a managed voice service like Vivox or Agora. Defer it until the core multiplayer loop is stable and validated with real players.

Have a project in mind? Contact Sodio Technologies to discuss your requirements and explore the right technology solution for your business.