Almost every agency answers this question with "it depends, contact us for a quote". It does depend — but that is not a reason to tell you nothing. You are trying to work out whether this project is a $3,000 idea or a $50,000 one before you waste anyone’s time, including your own. So here are real numbers, what moves them, and what the same app costs you every month after launch.
Those are our rates, and they are lower than you will be quoted by an agency in London or Austin for the same scope — we are based in Islamabad, so our cost base is different. The engineering is not different: the six streaming products we have shipped run on the same Agora, LiveKit and WebRTC stacks that everyone else uses. We quote in US dollars and can invoice in PKR for clients in Pakistan.
What actually drives the price
Before the tables, it is worth understanding what you are actually paying for, because two apps that look identical in a feature list can differ by a factor of five in build cost. Almost all of that difference comes from five things.
- Audio or video. Audio rooms are meaningfully cheaper to build and dramatically cheaper to run. If your market will accept voice-only, that single decision is the biggest saving available to you.
- Whether money moves through the app. A room where people chat is a small build. A room where coins are bought, converted to diamonds, split across four tiers and withdrawn to a bank account is a financial system that happens to have video in it.
- How many tiers the business has. Owner only is simple. Owner, BD, agency and host — each with its own panel, commission rate and view of the same ledger — is often a third of the total build.
- Games. Every in-room game is its own product with its own server, fairness model, wagering logic and anti-cheat layer. Games are usually priced separately for exactly this reason.
- How many edge cases you insist on at launch. Reconnects, background audio, low-end Android, weak networks, moderation and store review are where streaming projects actually spend their time. Cutting them does not save money, it defers it.
Cost to build, by scope
These are the four shapes almost every enquiry falls into. Ranges are for a complete, launchable product — apps, backend, admin, store submission — not a demo.
| Scope | Typical cost | Timeline | What you get |
|---|---|---|---|
| Audio room MVP | $1,500 – $4,000 | 4 – 6 weeks | Multi-seat voice rooms, mic queue, profiles, basic gifting and coins, simple admin panel, iOS and Android. |
| Live video streaming MVP | $4,000 – $9,000 | 8 – 12 weeks | Live video rooms with multi-guest, animated gifting, coin and diamond economy, in-app purchase, host earnings, moderation, admin panel. |
| Full Bigo-style platform | $9,000 – $25,000 | 4 – 7 months | Everything above plus PK battles, levels and badges, in-room games, fan clubs and rankings, plus owner, BD, agency and host panels over one ledger with payouts. |
| Multi-region / heavy games | From $25,000 | 6 months + | Multiple regional deployments, a full game suite, custom economy, localisation, and scale engineering. Scoped individually. |
The jump between the second and third row is where most budgets get a surprise. A video app with gifting is one product. A platform with a host hierarchy, commission rules and real payouts is three or four products sharing a database, and it is the tier structure — not the video — that accounts for most of the difference.
What in-app games add
Games are quoted separately because each one is genuinely its own build. A game that moves coins needs an authoritative server that decides every outcome, a provably fair random number generator, escrow and settlement so a round cannot pay out twice, and anti-collusion checks. The board is the cheap part.
| Game | Typical cost | Notes |
|---|---|---|
| Spin wheel / lucky box | $800 – $2,000 | Fast single-player rounds. Cheapest to build and usually the highest coin burn per user. |
| Dice, greedy & number games | $1,000 – $2,500 | Seconds per round, easy to understand, strong repeat spend in busy rooms. |
| Ludo (real-time multiplayer) | $2,500 – $6,000 | Turn timers, reconnects, auto-play on drop, spectators. The reconnect handling is most of the work. |
| Teen Patti / card games | $3,000 – $7,000 | Server-dealt cards, pot management, side-show rules, and collusion detection across linked accounts. |
| Fishing / slots / arcade | $2,000 – $5,000 | Per-title RTP configuration and heavier art requirements. |
| Full game suite (4 – 6 titles) | $8,000 – $15,000 | Cheaper together than separately, because the fairness, wagering and settlement layer is built once and reused. |
What the admin, agency and BD panels add
This is the line item most quotes leave out entirely, and the one that causes the most trouble later. If your business model involves agencies recruiting hosts, you need this — and bolting it on after launch costs more than building it in, because by then the coin records already disagree with each other.
| Scope | Typical cost | What it covers |
|---|---|---|
| Admin panel only | $1,200 – $3,000 | Users, rooms, reports, bans, coin and gift catalogue, basic revenue dashboard. |
| Owner + agency panels | $2,500 – $6,000 | Agency accounts, host rosters, targets, agency earnings and withdrawal requests. |
| Full owner / BD / agency / host | $5,000 – $12,000 | Four tiers over one double-entry ledger, configurable commission engine, payout approval workflow, fraud controls and reconciliation reporting. |
What it costs to run every month
Build cost is a one-off. Running cost is forever, and it is the number that decides whether the business works. We are deliberately not quoting vendor rate cards here — they change, they vary by region and by negotiated contract, and a blog post repeating last year’s prices is worse than no post. What follows is the shape of the bill and which lines actually matter.
- Streaming minutes, if you use a managed SDK. Billed per participant-minute, so a room with one host and fifty viewers bills fifty-one people per minute, not one. This is almost always the largest line and it grows with every successful month.
- Servers, if you self-host. A LiveKit deployment replaces per-minute billing with fixed instance cost — predictable, and it does not rise just because your app had a good week.
- Bandwidth. The real cost of self-hosting, and the per-gigabyte rate varies enormously between providers. Pick the wrong host and it erodes the saving.
- Backend, database, Redis and object storage. Modest at early scale and easy to forecast.
- Store fees. Apple charges $99 per year for the developer program; Google Play is a one-time $25. Both take a commission on in-app purchases, which is a revenue line, not a hosting line.
- Maintenance. OS upgrades, SDK updates, store policy changes and bug fixes. Budget for it as an ongoing retainer rather than pretending launch is the end.
The single biggest lever on running cost is whether you stay on a per-minute SDK or move to self-hosted infrastructure. At prototype scale a managed SDK is clearly right. Once rooms are consistently busy, per-minute billing scales against you precisely when you are winning. We have written about that trade in detail, including where self-hosting is the wrong answer.
Why the same brief gets quoted at $2,000 and $80,000
If you have sent the same specification to five teams you have probably had five wildly different numbers back. There are only a few explanations, and it is worth knowing which one you are looking at.
- It is a clone script with your logo on it. Very cheap, running in days, and usually encrypted code plus a licence fee, a revenue share, or a shared streaming account you do not control. Fine for testing a market; a problem the moment you need a BD tier or real moderation.
- The quote silently excludes the panels. A quote for the apps only, with agency and payout tooling left out, can look half the price of a complete one. Check what is in scope before you compare numbers.
- The quote silently excludes the economy. Coins in and diamonds out is where the engineering actually is. If nobody asked you about commission splits, they have not priced them.
- Rate base. A team in Western Europe or North America has a cost base several times ours. That is a real difference in price, not in competence — and it cuts the other way too, so ask who is actually writing the code.
- Someone is guessing. If a fixed price arrives without anyone asking about your average room size, your regions or your payout model, it is a number, not an estimate.
How to spend less without wrecking the product
There are good ways to cut a streaming budget and bad ways. These are the good ones, in the order we usually recommend them.
- Launch audio-only. Cheaper to build, dramatically cheaper to run, and in several of the biggest markets for this category it is what users actually prefer. Video can come in phase two.
- Ship one game, not six. Pick the fast one — a wheel or a lucky box. It will earn more than a half-finished Ludo, and the fairness and settlement layer it establishes makes the next game much cheaper.
- Start with owner and agency tiers only. Add the BD layer when you actually have regional partners, provided the ledger was designed to expect it.
- Launch in one region. Multi-region deployment multiplies both infrastructure and testing. Prove the model in one market first.
- Do not cut moderation. It is the one saving that reliably costs more than it saves, because the app gets rejected by Apple or Google and you pay for it twice.
- Do not cut the ledger. If coins are stored as a single balance field rather than a ledger of entries, your payouts will stop matching and no amount of later work fixes it cleanly.
What we charge, and why it is lower
We are a Pakistan-based team and most of our clients are in Pakistan, the Gulf, the United Kingdom and the United States. Our rates reflect where we are, not what we can build: the same six shipped streaming products, the same Agora, LiveKit and WebRTC stacks, the same Flutter and Node.js codebase quality that an agency charging four times as much would deliver.
What we will not do is quote you a number before we understand the model. Every quote we send follows a free call where we ask about your expected room sizes, your regions, your commission structure and how money comes in and goes out — because those four answers determine the price far more than the feature list does. The quote that follows is fixed scope, in writing, with a timeline.
If a quote arrives without anyone asking how your coins turn into cash, it is not an estimate. It is a guess you will pay for twice.
The bottom line
Budget $1,500 to $4,000 to find out whether your idea has legs as an audio room, $4,000 to $9,000 for a credible video streaming product with a working economy, and $9,000 to $25,000 for the full Bigo-style platform with PK battles, games and the agency hierarchy behind them. Add games and panels on top of that, and plan for monthly running costs that are dominated by either streaming minutes or servers, depending on which way you go.
Send us what you have — even if it is a paragraph and a link to an app you like — and we will come back with a scoped number rather than a range. If your budget does not fit the product you are describing, we will tell you that too, and suggest the smaller version worth building first.