Mobile app development cost explained: 2026 ranges from five sources, what drives them, store fees, maintenance at 15-20% a year and how to read quotes.
Quick answer: Mobile app development cost depends on a few decisions rather than one headline number: how many platforms you target, how many features and integrations you need, how much backend must be built, and how much the app needs after launch. Published ranges for similar apps differ several times over, so define scope first, then compare quotes.
Nobody publishes a standard price for an app, because the price is developer time, and developer time depends on scope. Two quotes for what sounds like the same app can differ several times over, so a single headline number says little about your project.
A better way to think about it: cost is roughly the hours your scope needs multiplied by the hourly rate of the team doing it, plus the running costs that continue after launch. Scope moves the hours; team location and seniority move the rate. The rest of this guide takes each lever in turn so you can see what you are paying for.
Singapore teams generally charge more per hour than teams in lower-cost regions, and Singapore buyers often also pay for account management and local compliance knowledge. The practical advice is the same anywhere: write the scope down, ask several providers to price that same scope, and compare the totals and exclusions rather than the hourly rates. A cheaper hour that produces more rework is not cheaper.
Developer time is the cost; these are the levers that buy more or less of it.
| Driver | How it moves cost | How to control it |
|---|---|---|
| Platforms (iOS, Android, web) | Each separate native codebase is a second build and a second test cycle. | Launch on the platform your first users actually carry, or use one shared codebase. |
| Feature count and complexity | A simple login is small; payments, chat, offline mode and real-time tracking are each a substantial piece of work. | Rank features; ship only the must-haves first. |
| Backend and admin tools | Most business apps are mostly backend: accounts, data, permissions, notifications and an admin panel. None of it shows in a screenshot. | Ask for the data model and admin screens in the estimate, not as an afterthought. |
| Third-party integrations | Payment gateways, maps, CRMs, SMS and WhatsApp each bring documentation to read, edge cases to handle and often usage fees. | List every system the app must talk to before you ask for a price. |
| Design depth | Custom animation and bespoke components cost more than standard platform patterns. | Use standard components first; spend design effort on the two or three key flows. |
| Compliance and security | Handling personal data, payments or health information adds review, testing and documentation. | Name the applicable rules up front (see the compliance section below). |
| Team location and seniority | Hourly rates differ several times over between regions; senior time costs more but often wastes less. | Compare the total for a defined scope, not the hourly rate alone. |
| Ongoing maintenance | OS and store rules change every year, so an unmaintained app eventually stops being accepted. | Agree a maintenance plan before launch. |
Drivers of mobile app cost and how to manage each.
Design is its own line in a good estimate. Our UI/UX design service lets teams test flows cheaply before a developer writes any code.
Cross-platform is usually cheaper than two native apps, but how much cheaper depends on the product.
The saving is real when the two apps are largely the same and shrinks when a product depends on platform-specific features. On the framework side, Flutter says you can build mobile, web and desktop apps from a single codebase, and React Native says its components map directly to native UI building blocks and that you can write your own native code where needed.
| Native (Swift and Kotlin) | Cross-platform (Flutter or React Native) | Web app or PWA | |
|---|---|---|---|
| Codebases | Two | One shared, plus platform-specific code where needed | One, runs in the browser |
| Cost position | Highest of the three, because two codebases are built, tested and kept in step | Usually lower than two native apps; the saving depends on how alike the two apps are | Usually lowest to build |
| Best for | Heavy camera, AR, background processing, newest OS features | Most business, booking, ordering and marketplace apps | Content, dashboards and internal tools where a store listing is not needed |
| Store presence and device access | Full | Full, with native modules for any gaps | Limited |
| Main risk | Cost, and keeping two apps in step | Native-module gaps and framework upgrades | Apple rejects apps that are only repackaged websites |
Decision matrix. Cost positions are our summary; the framework facts come from Flutter and React Native documentation.
For a specific product, our iOS, Android and cross-platform app development page explains how we approach the choice, and a custom web application is often the cheapest way to test an idea before committing to the stores.
The fees are small; the rules cost time. Apple's Developer Program is US$99 a year (the Enterprise Program is US$299). Google charges a one-time US$25 registration fee. Google also requires new personal developer accounts to complete a closed test before launch, with a minimum of 12 testers opted in continuously for at least 14 days. That is a schedule item, easy to miss when working back from a launch date.
Commission applies to apps selling digital goods. Apple's Small Business Program lowers its commission to 15 percent for developers with up to US$1 million in proceeds in the prior year. Google's service fee page sets 15 percent on the first US$1 million of annual revenue and on automatically renewing subscriptions, and 30 percent on earnings above US$1 million in that tier.
Review rules also shape the build. Apple's App Review Guidelines say an app should go beyond a repackaged website, that an app supporting account creation must offer account deletion within the app, and that features such as subscriptions and premium content must use in-app purchase. Every app needs a privacy policy linked in the store listing and inside the app. Building these in early is cheap; retrofitting after a rejection is not.
Wherever the app handles personal data. In Singapore, the maximum financial penalty under the Personal Data Protection Act is S$1 million, or 10 percent of annual Singapore turnover for organisations above S$10 million in turnover, for contraventions from 1 October 2022, according to law firm Allen & Gledhill.
Expect more testing, a data-retention design and a deletion path. If your app touches regulated data, involve security and compliance specialists before the design is final, and read our guide to data privacy for businesses for the broader picture.
Because the platforms keep moving. Google Play requires new apps and updates to target Android 16 (API level 36) or higher, with a deadline of 31 August 2026 and an extension available to 1 November 2026. Apple requires uploads to be built with Xcode 26 or later using an SDK for iOS 26 (since 28 April 2026) and to target iOS 13 or later from 9 September 2026. An app nobody touches will eventually stop being accepted for updates.
Budget for maintenance as a recurring line from the first year, not as a surprise. Usage-based services add more: Google Maps Platform, for example, bills per call beyond a free monthly allowance.
This is why we treat application maintenance and support as part of the plan rather than an optional extra.
Three lessons from software we build and run, all about cost that arrives after the feature list is finished.
Credentials quietly expire. Our daily Google Ads monitor went dark mid-day because Google expires OAuth refresh tokens after seven days when the OAuth app is left in "Testing" status. The durable fix was a service account added to the ad accounts as a user, which never expires. Nothing in the original feature list mentioned token lifetimes, and an estimate that stops at "integrate with the API" will not either.
The failure is often in the plumbing. A hosted chatbot we run went dark because of a CDN security rule and an expired credential, not because of anything in the application logic. That is why we monitor DNS and CDN rules, tokens and database quotas, not just the app itself. Budget for monitoring the same way.
A successful upload is not a successful deploy. On shared hosting the FTP root is not always the site's web root, and an upload to a wrong-but-plausible folder succeeds silently. Re-downloading the file you uploaded proves nothing, so we verify every deploy over the public HTTPS URL. Release process is real work, and a quote that includes it is a better quote.
On one client tool, a report filter was called broken three times. We fixed four genuine interface bugs before finding the real cause: a missing Cache-Control header. Maintenance often starts with diagnosis, not code.
Our published work shows the same pattern at scale. The on-demand services app and the cab booking app case studies describe app builds; the SaaS billing platform (40 percent fewer billing support tickets) and the inventory and order management system (stockouts down 60 percent) show what backend and integration work delivers. Related builds sit under software and app development, and SaaS product development.
Because the biggest risk is building the wrong thing. CB Insights' analysis of 431 venture-backed companies that shut down since 2023, published in March 2026, found poor product-market fit cited in 43 percent of cases, and running out of capital in 70 percent, which the report describes as usually the final cause rather than the root problem.
An MVP is not a cheap version of everything. It is the smallest release that tests the one assumption your business depends on. A practical sequence:
If you are a founder, our software development for startups approach is built around this order of work, and a discovery call can help you decide what to cut. If a workflow fix would do, our guide to business process automation compares the options.
The more specific your brief, the narrower and more honest the range you get back. Prepare these:
Check the pricing model, then the exclusions.
| Model | How it works | Watch for |
|---|---|---|
| Fixed price | One price for a defined scope, often paid in milestones. | Change requests are priced separately, so the written scope matters more than the headline figure. |
| Time and materials | You pay for hours worked, usually with an estimate and a cap. | Without a cap and regular reporting, costs can drift. Ask for weekly burn reports. |
| Phased or retainer | A discovery phase first, then build and support in stages. | Make sure each phase has a deliverable you can accept or walk away from. |
Common pricing models for app projects. Scope and price at DXDasia are confirmed in writing after a discovery call.
Then ask what is not included. Common gaps are: app store accounts and fees, hosting and cloud, third-party service fees, quality assurance and testing time, compliance work, project management, content and translation, analytics set-up, and anything after launch.
Across several client accounts we have seen a website redesign remove the Google Ads conversion tag, so the account kept spending while reporting zero or wrong conversions. If your app or site is rebuilt, ask who re-verifies the tracking afterwards, and whether that is in the quote.
Do not choose a vendor on the headline figure. The published ranges overlap so heavily that price alone tells you almost nothing, while scope, exclusions and the plan for after launch tell you a lot. Our recommendation: start with an MVP on one shared codebase unless your product clearly needs native-only features, insist on a written list of exclusions, and budget for maintenance as a recurring annual cost. Treat any quote that cannot explain its backend, integration and testing lines as incomplete. If you would like a scoped view of your own idea, we are happy to talk through it on a discovery call; scope and price are confirmed in writing afterwards. Our articles on building an AI agent and AI chatbot costs cover adjacent builds.
There is no standard price. Cost is developer hours multiplied by the team's hourly rate, plus running costs after launch, so it rises with the number of platforms, features, integrations and backend work. Two quotes for the same idea can differ several times over. Define your scope first, ask several providers to price it, and compare totals and exclusions rather than headline numbers.
Usually, because one shared codebase replaces two separate builds, but the saving depends on the product. It is largest when both platforms share most features and smallest when the product relies on platform-specific capabilities such as AR or heavy camera use. Ask each provider to price both routes for your scope.
Treat maintenance as a recurring annual cost, not a one-off. It covers OS and store-rule updates, hosting and cloud services, third-party fees and bug fixes. Platforms change their requirements every year, so an app nobody maintains eventually stops being accepted for updates. Agree the maintenance plan before launch.
Yes. Apple's Developer Program costs US$99 a year, and Google Play charges a one-time US$25 registration fee. Both also take a commission on digital sales: Apple's Small Business Program and Google's reduced tier set 15 percent on the first US$1 million a year. New personal Google accounts must also complete a 14-day closed test with 12 testers.
In most cases, yes. An MVP tests the one assumption your business depends on before you spend on everything else. CB Insights found poor product-market fit cited in 43 percent of 431 startup shutdowns since 2023. Keep the first release to the features that solve one problem for one type of user.
Fixed price suits a well-defined scope but treats changes as extra cost, so the written scope matters. Time and materials suits evolving products but needs a cap and regular reporting. A phased approach, with discovery first, often gives the best control. Whichever you choose, ask for the exclusions in writing.
Common exclusions are app store accounts, hosting and cloud fees, third-party service fees, testing time, compliance work, content and translation, analytics set-up and post-launch support. Check each one against your list. A quote that does not mention QA or backend work has probably left them out rather than included them.