The honest answer to what a mobile app costs is a range, and anyone who gives you a single number before understanding your product is guessing. But the ranges are knowable, and knowing them protects you from both the $8,000 quote that ends in an unusable codebase and the $400,000 quote for something that should have cost a quarter of that.
For a production-quality MVP from a credible agency in 2026, expect $40,000 to $150,000. That buys one or two platforms, a working backend, authentication, a real design pass, testing, and store submission. Below roughly $25,000 you are buying either a template, a very junior team, or a project that will need rebuilding within a year. Above $150,000 you should be getting genuine complexity — real-time infrastructure, regulated payments, machine learning, or a multi-sided marketplace.
Platform choice is the first big lever. Building native iOS and Android separately roughly doubles the client-side work. Cross-platform with React Native or Flutter typically lands 30–40% cheaper for the same feature set and is the right default unless you need heavy device-level performance, deep OS integration, or advanced graphics. A simulator game or an AR product should be native or engine-based; a marketplace, a health tracker, or a fintech dashboard almost never needs to be.
Backend complexity is the biggest hidden multiplier. An app that reads from a content API is cheap. An app with user-generated content, payments, real-time sync, notifications, moderation, and an admin panel is a different product entirely — the visible screens might be identical, but the engineering underneath is three times the work. When comparing quotes, compare backend scope line by line, because that is where two proposals for the same app diverge.
Payments and compliance deserve their own budget line. Taking card payments is a solved problem until you need marketplace payouts, multi-currency, refunds, chargeback handling, or custody of anything. Regulated sectors — fintech, health, anything touching children's data — add legal review, audit trails, data residency, and security work that can add 20–40% to a build. Budget for it up front rather than discovering it in month four.
Design is not a line item you should cut. A proper product design phase — user flows, a component system, prototypes, and a design QA pass at the end — usually runs 15–25% of the build. Skipping it does not save money; it moves the cost into rework, because the team ends up designing in code, badly, under deadline.
Post-launch is the cost founders consistently underestimate. Plan for 15–20% of the original build cost per year just to stand still: OS updates, dependency upgrades, crash fixes, store policy changes, and server costs. That is maintenance, not new features. If you intend to keep shipping, budget a retainer or an in-house engineer on top.
The cheapest real way to reduce cost is to cut scope, not quality. Ship one platform first. Ship three features that work instead of ten that half-work. Use managed infrastructure instead of building your own. Reuse a proven component library rather than designing every control from scratch. Every one of those decisions is reversible later, whereas a cheap codebase is not.
At Brainbaby Labs we build custom mobile apps across fintech, health, marketplaces, and AI — and we scope honestly before quoting. If you want a realistic number for your project rather than a sales figure, get in touch.