What is actually left over

Revenue is the easy half. This page subtracts the app stores, the creators, the payment rails, the streaming infrastructure and the people, and shows what remains at any scale you choose.

These are assumptions, not measurements

Nexora has not launched, so there is no observed ARPU, no measured churn and no real infrastructure bill. Every number below is an input you set, and the defaults are deliberately conservative rather than flattering. Treat the output as a model of your assumptions — it tells you which assumption matters most, not what will happen.

Demand

Monthly active users—

Share who ever pay—

Live-gifting platforms typically convert 1–5% of monthly actives into payers. Below 1% no amount of ARPU saves the model.

Monthly spend per payer—

Heavily skewed in practice — a handful of whales carry the average. The mean is what the model needs.

Bought via web checkout—

The rest goes through Apple or Google at 30%. This one slider moves profit more than any other on the page.

Revenue mix

Gifting—

Subscriptions—

Commerce & products—

—

Blended creator split—

Applied to the bean pool on gifting and to net receipts on subscriptions. Commerce and digital products use a flat 12% marketplace commission instead, so this slider does not move them.

Cost base

Streaming architecture

Illustrative rates. Replace with your negotiated pricing before relying on the output.

Live minutes watched per user—

Per month, averaged across all monthly actives — not just viewers of live.

Free AI answers per user—

Live translation, captions and dubbing — given away, not billed. Paid AI twin questions are modelled on the streams page where they carry their own revenue. At 3¢ each this is the only cost that scales with usage rather than with users, and the only one that can outrun revenue.

Team & overheads—

Monthly. Salaries, moderation, support, legal, marketing.

💵

Monthly gross revenue

—

📈

Monthly net profit

—

Where every dollar goes

—

Contribution per user

Revenue / user

—

Variable cost / user

—

Contribution / user

—

The same assumptions at four scales

Users	Gross	Creators	Costs	Profit	Margin

Team cost is held flat across all four rows, which is why margin appears to improve with scale. In reality moderation and support headcount grow with users — the improvement is real but smaller than this table implies.

What actually moves the number

Each row shows the monthly profit change from moving one input, holding everything else at your current settings. Ranked by impact.

What this model omits

📉

User acquisition costAssumed zero. Paid installs in live-streaming categories are not cheap, and they land entirely against the payer conversion rate above.

🎬

Seed creator guaranteesPaying 50 creators to go live daily is a real pre-launch line item and it is not in the cost base.

⚖️

Chargebacks and fraudThe 5% reserve covers them if it is funded. Unfunded, they hit profit directly.

🌍

Tax and withholdingUS withholding, EU VAT and local creator tax obligations all change the net payout, not the gross.

The three decisions that matter

1 · Get buyers onto web checkout

Thirty percent versus three percent, on the largest stream. Nothing else on this page is worth as much, and it requires no change to what creators are paid.

2 · Cap AI cost per interaction before launch

It is the only cost that scales with usage rather than users, and it is the only one that can make an individual transaction unprofitable. A token ceiling costs a day to build.

3 · Pick the split model once, then leave it

Moving the blended split by ten points changes profit less than the store-cut decision, and cutting it after launch is the fastest way to lose the creators the pitch attracted.