TL;DR — cost at a glance
Building a neobank MVP on a banking-as-a-service platform runs $250,000–$500,000. A fuller production version lands at $500,000–$1,000,000+. The trick that makes those numbers possible: you rent the bank instead of building it. A sponsor bank holds the licence while you build the apps, onboarding/KYC, accounts and cards. What eats the budget is compliance and the banking core, not the screens people tap. And whether the whole thing survives comes down to per-account run-cost.
Most founders overestimate what a neobank costs to build. The reason is simple. The expensive part — the bank itself — gets rented, so the build budget covers only the product around it. Here is what that leaves you paying for:
- MVP cost: $250,000–$500,000 on a banking-as-a-service (BaaS) platform — apps, onboarding/KYC, accounts, a debit card and core payments.
- Production neobank: $500,000–$1,000,000+ with lending, multiple products and a richer feature set.
- You don't build the bank: a sponsor bank holds the licence and the deposits through a BaaS provider, and you build the product on top.
- Cost is in compliance and the banking core: KYC/AML, accounts and ledger, card issuing. Not the app screens.
- Run-cost matters more than build-cost: the per-account BaaS, KYC and card fees are what decide whether the unit economics work.
- Phase the MVP: one account, one card and core payments come first, with lending and extras left for later.
What is a neobank — and what do you actually build?
A neobank is a digital-first bank with no branches, delivered through a mobile app. But "building a neobank" rarely means building a bank. In practice you build four layers on top of a banking partner:
- Customer apps — iOS, Android, sometimes web. The experience customers see.
- Onboarding & identity — KYC/AML, document and identity verification, sanctions screening.
- Banking core — accounts, balances, a transaction ledger, statements. Usually provided by the BaaS partner, orchestrated by your backend.
- Card program — issuing virtual and physical debit cards, spend controls, real-time transaction notifications.
Fraud monitoring, support tooling, notifications and analytics sit around those four layers, and the budget spreads across all of them. Our fintech industry page shows how the pieces fit together.
Examples of successful neobanks
The neobanks that reached scale almost all launched narrow: one product on a partner bank, widened only after demand showed up. Copy that phasing when you cost your own launch. A big-bang full-bank build is the opposite of what these companies actually did.
| Neobank | Market & launch | How it started |
|---|---|---|
| Revolut | UK/EU, 2015 | A multi-currency card and FX app; accounts, stocks and crypto came later |
| Chime | US, 2013 | A fee-free spending account and debit card on a sponsor-bank model |
| N26 | Germany/EU, 2013 | A mobile current account and card first; its own banking licence came years later |
| Nubank | Brazil, 2013 | A single no-fee credit card before expanding into a full digital bank |
| Monzo | UK, 2015 | A prepaid card MVP first, then a full current account once product-market fit was proven |
The cost lesson runs through all five. Each started with a single product on rented banking rails, then expanded once demand was proven. Budget for that first slice. The whole bank comes later, if it comes at all.
Build vs banking-as-a-service
This one decision sets the cost. You have two routes.
Banking-as-a-service (BaaS) — you build on a provider that supplies accounts, cards and a sponsor-bank relationship via API (Unit, Treasury Prime, Stripe Treasury in the US; Solaris, Swan in the EU). You own the product. The bank owns the charter, the deposits and the core regulatory burden. Almost every modern neobank launches this way, because it is the fastest and cheapest compliant path there is.
Building your own banking core / pursuing a licence — multi-year, multi-million-dollar work that only makes sense at significant scale, once the BaaS fees finally outweigh the cost of owning the stack. For a launch, it is the wrong route, full stop.
Cost breakdown by module
Indicative build costs for a BaaS-based neobank MVP, by module. Ranges vary with scope, market and partner.
| Module | Build cost | Notes |
|---|---|---|
| Mobile apps (iOS + Android) | $70k–$140k | Cross-platform or native; the visible product |
| Onboarding + KYC/AML | $40k–$90k | Vendor integration + flows; per-check fees separate |
| Accounts, ledger & BaaS integration | $60k–$130k | Orchestration, reconciliation, statements |
| Card program (issuing, controls) | $40k–$90k | Virtual + physical, notifications, limits |
| Payments (top-up, transfer, spend) | $30k–$70k | Rails depend on market (ACH/SEPA) |
| Support, fraud, ops tooling | $30k–$80k | The layer teams routinely under-budget |
Add those up and an MVP lands in the $250,000–$500,000 range. Want to cost the app layer on its own? Our mobile app development cost guide breaks it down, and our payment gateway integration guide covers the payment specifics.
Licensing and the sponsor-bank reality
In both the US and the EU, moving money and holding deposits requires a licence. The sponsor bank holds it, not you. Working through a BaaS provider, the chartered bank carries the regulatory obligation while you operate the brand and the product. That one arrangement separates launching in months from launching in years. It also explains why neobank build cost is measured in hundreds of thousands rather than the tens of millions a charter would imply. Our fintech app development guide covers the deeper compliance backdrop.
Compliance and run-cost
Two cost streams continue after launch and decide viability:
- Compliance: KYC/AML verification and monitoring, transaction screening, dispute and chargeback handling, and the controls the sponsor bank requires. Mostly per-user and ongoing.
- Platform run-cost: BaaS platform and per-account/per-card fees, card scheme and processing fees, cloud infrastructure, and support.
Model both per user before you build anything. A neobank that wins customers but loses money on every account ends up worse off than one that scoped carefully and proved the spread first.
How do neobanks make money?
Build cost is modest. Per-account run-cost is not, so the revenue model is what decides whether a neobank survives. These are the main streams:
- Interchange — a share of the fee merchants pay on every card transaction; the primary revenue for most consumer neobanks.
- Subscription tiers — paid plans (premium cards, insurance, higher limits, FX perks) that lift average revenue per user.
- Lending & credit — overdrafts, credit lines and BNPL, usually added after launch once the base product works.
- Interest & treasury — earnings on the deposits held with the partner bank.
- FX & cross-border fees — margins on currency conversion and international transfers.
- Marketplace & referral — commissions from third-party financial products offered in-app.
Interchange alone rarely covers the fully-loaded cost of an active user. That is precisely why the unit-economics modelling above matters so much. Pick your revenue streams while you are still scoping, well before the first sprint.
Timeline, team and MVP phasing
A BaaS-based neobank MVP typically takes 6–9 months. Here is the part teams miss: BaaS and sponsor-bank onboarding, KYC vendor contracting and program approval all run in parallel with engineering, and they are often the real critical path. Start them in week one. A typical team looks like this — a product/delivery lead, mobile engineers, two backend engineers (one owning the ledger and integration), a QA engineer with security skills, plus part-time DevOps and compliance input.
Phase the MVP. Ship onboarding, one account, a debit card and core payments first. Push lending, savings and interest, multi-currency, business accounts and rich personal-finance features to later releases. That discipline keeps the first build in the $250,000–$400,000 range and gets you to market, and to real data, faster. Many fintechs assemble the team through a dedicated development team to keep the cost in check.
How to control the cost
- Launch on BaaS. For an MVP, never build the bank yourself.
- Scope a tight MVP: one account, one card and core payments, with everything else deferred.
- Model run-cost first. Unit economics decide viability far more than the build budget does.
- Start partner and compliance onboarding in week one. That work, not the code, is usually the critical path.
- Choose a partner who has shipped a regulated fintech before. The operational layer is exactly where inexperience gets expensive.
At its heart this is mobile and custom software work. The right team and a phased plan remain the two levers that move the cost most.
FAQ
How much does it cost to build a neobank in 2026?
Budget $250,000–$500,000 for an MVP on a banking-as-a-service platform. That covers the apps, onboarding/KYC, accounts, a debit card and core payments. A broader production neobank with lending and multiple products pushes past $500,000 and can top $1,000,000. Most of that money goes to KYC/AML, the banking core and compliance rather than the app screens. And ongoing BaaS and compliance fees sit on top, separately.
Do I need a banking licence?
Almost certainly not. Most neobanks run on a sponsor-bank model through a BaaS provider, where the chartered bank holds the licence and the deposits while you own the product. Chasing your own charter is a multi-year, multi-million-dollar undertaking, and it only starts to make sense at significant scale.
What is the cheapest way to launch?
Build on a BaaS provider (Unit, Treasury Prime, Stripe Treasury in the US; Solaris, Swan in the EU) and scope a tight MVP: onboarding, one account, a debit card and core payments. That route typically lands at $250,000–$400,000, and it is the fastest compliant path to launch.
What drives most of the cost?
Compliance and the banking core, not the UI. KYC/AML, the accounts and ledger layer, card issuing, transaction monitoring and the controls the sponsor bank requires are where the money goes. Two apps with identical screens can differ several-fold in cost, and the difference is everything sitting behind a neobank's screens.
Can a startup afford it?
Yes, with a tight MVP on BaaS and a cost-efficient engineering partner. The bigger risk was never the build budget. It is unit economics, so prove the spread between revenue and per-user cost before you scale features.
Last updated 3 July 2026. Cost ranges reflect BaaS-based agency builds for US and EU markets and vary by scope, products, market and partner. Regulatory references are general guidance, not legal advice — consult qualified counsel for your jurisdiction. Request a scoped proposal for your specific neobank.
