Cost to build a React Native mobile app in 2026
Cost to build a React Native mobile app in 2026
Building a React Native mobile app in 2026 typically costs $20,000 to $150,000 to ship a real V1, depending on whether you stay inside Expo's managed workflow or escape to bare React Native with custom native modules. The biggest cost drivers most articles skip: native module work, EAS subscription fees, and whether your team picked the right starting workflow in week one.
Most "React Native cost" posts you'll find on Google are agency landing pages: regional rate table, feature cost grid, "contact us" button. They don't tell you that picking Expo vs bare RN can swing your build by $10,000, or that a single CallKit-style native module can eat two weeks. This post does.
What you actually pay for in a React Native app
Costs cluster into five buckets. The first one swallows the rest.
- Engineering hours. 70 to 85% of your spend. A 12-week V1 with one Mid-level engineer is typically $12,000 to $25,000 of pure dev cost, depending on how you book.
- Apple Developer + Google Play. $99/year for Apple, $25 one-time for Google. Non-negotiable.
- Expo EAS subscription. Free tier covers small projects. Production tier is $99/month and unlocks priority builds, larger build resources, and EAS Update at scale.
- Backend SaaS. Supabase, Clerk, Stripe, Sendbird, OneSignal. Most apps run $200 to $1,500/month in V1.
- Design + QA. $3,000 to $15,000 if outsourced. $0 if you have a designer or use Figma templates.
The number you see quoted on agency sites ($60k, $90k, $180k) is mostly the first bucket with a 2 to 3x markup. If you book engineering directly, the rest of the math gets a lot kinder.
Expo vs bare React Native: the cost fork most articles skip
This is the single biggest decision affecting your total. Every article on page one of Google for "cost to build react native app" pretends the choice doesn't exist.
Expo Managed Workflow in 2026 is the default for most teams. You get:
- EAS Build (cloud builds for iOS and Android, no Mac required)
- EAS Submit (push to App Store + Play Store from one command)
- EAS Update (over-the-air JS updates without resubmitting)
- Config plugins for most popular native modules (Stripe, Notifee, RevenueCat, react-native-vision-camera)
- A free dev client that lets you test pushes, deep links, and biometrics on real devices
Bare React Native gives you full native access. You can write Swift and Kotlin, drop in any third-party SDK, modify the iOS/Android project files directly. You can also break things in ways Expo would have prevented.
Here's the cost difference nobody puts in a table:
| Workflow | Bootstrap | Native module work | Build infra | OTA updates | Best for |
|---|---|---|---|---|---|
| Expo Managed | 2-4 hours | $0 if config plugin exists | EAS handles it | EAS Update, included | 80% of apps in 2026 |
| Bare RN | 1-3 days | $2k-8k per module | Self-hosted CI or EAS Bare | CodePush or EAS Update | Apps needing CallKit, custom Bluetooth, exotic hardware |
A team that starts bare "for flexibility" and never needs it wastes $8k to $15k on bootstrap, build infra, and dependency upgrade pain over the first year. A team that starts in Expo and has to eject mid-project pays a one-time $3k to $6k tax. Net: start in Expo unless you know on day one you need a native module Expo can't host.
Real "must escape Expo" cases in 2026:
- VoIP apps using CallKit / ConnectionService end-to-end
- Bluetooth Classic (Expo handles BLE fine)
- A specific OEM SDK with no config plugin and no maintainer
- Heavy custom video pipelines (some, not all)
If you're building a SaaS dashboard, a marketplace, a delivery app, or anything you'd compare to a marketplace V1 build, Expo Managed will carry you to launch.
Cost breakdown by approach
Here's what the same 12-week React Native V1 costs across the realistic options in 2026.
| Approach | Cost | Timeline | Pros | Cons |
|---|---|---|---|---|
| US full-time RN hire | $160k-220k/yr + benefits + equity | 3-6 weeks to hire, 8-16 weeks to ship V1 | Deep ownership, in-meeting, long-term roadmap | Slow to start, can't replace easily, expensive on burn |
| RN dev agency (US/EU) | $60k-180k for V1 | 10-20 weeks | PM included, design bundled, SOW-driven | 2-3x engineering markup, change orders, you don't pick the engineer |
| Upwork freelancer | $15k-50k | 12-24 weeks | Cheap, large pool, async-friendly | Vetting variance, no replacement guarantee, ghosting risk |
| Toptal / Lemon.io | $80-150/hr | 1-2 weeks to match, 8-16 weeks to ship | Vetted seniors, faster than full-time hire | Hourly billing creep, monthly contracts, recruiter call required |
| Cadence | $500-$2,000/week | 48-hour trial, ships from week 1 | Every engineer AI-native by default, weekly billing, replace any week, no notice period | Less suited to enterprise procurement, no fixed-bid contracts |
A few honest notes on this table:
- A US full-time RN engineer is the right call if you're past product-market fit and want a long-term mobile lead. It is not the right call for a V1 you're not sure will ship.
- An agency is worth the markup when you need a PM + designer + QA in one shop and can't run a vendor yourself. Otherwise you're paying for overhead that doesn't ship code.
- Toptal wins on perception if your CTO needs to point at a logo. It loses on price and contract flexibility versus newer marketplaces; Toptal alternatives for startups covers the trade-offs.
- Cadence is an on-demand engineering marketplace where founders book engineers by the week. Every engineer on Cadence is AI-native, vetted on Cursor, Claude Code, and Copilot fluency before they unlock bookings. A Mid at $1,000/week for 12 weeks is $12,000 of engineering for a real V1, with a 48-hour free trial and the option to replace anyone next Monday.
Feature-by-feature cost in React Native specifically
The cost depends heavily on what you buy versus build. Here's the 2026 RN-specific math, with real SDKs and real prices.
| Feature | Build hours | SaaS option | SaaS price | RN-specific notes |
|---|---|---|---|---|
| Email + social auth | 8-16 hr | Clerk Expo SDK | Free up to 10k MAU, $25/mo+ | Clerk's Expo support is genuinely first-class in 2026 |
| Payments | 12-24 hr | Stripe React Native SDK | 2.9% + 30¢ | PaymentSheet handles Apple Pay + Google Pay out of the box |
| Push notifications | 4-12 hr | Expo Notifications (free) or OneSignal ($9+) | Free or $9/mo | Expo Notifications free until you need segmentation |
| In-app chat | 24-60 hr | Stream Chat / Sendbird | $499/mo+ for production | Building chat from scratch is the most under-estimated feature |
| Maps + location | 8-20 hr | react-native-maps + Google Maps API | Pay-per-request | iOS uses Apple Maps, Android uses Google. Plan for both. |
| Video playback | 12-30 hr | Mux ($50/mo+) or Cloudflare Stream | Pay-per-minute | expo-video shipped stable in 2024 and is solid in 2026 |
| Heavy animations | 16-40 hr | Reanimated 3 + Skia | Free (libraries) | This is where RN's perf ceiling has moved most since 2022 |
| Analytics | 4-8 hr | PostHog / Amplitude / Mixpanel | Free tier, $20/mo+ | Use the SDK, don't roll your own |
| AI features | 16-60 hr | OpenAI / Anthropic / Vercel AI SDK | Per-token | If you're adding chat-style AI, the cost of AI chatbot integration maps almost 1:1 to RN apps |
Two things this table makes obvious. First, if your "mobile app" is mostly screens, lists, and CRUD, you can buy 60% of the work and ship in 6 to 10 weeks. Second, the moment you need real-time anything (chat, video calls, live tracking), costs jump fast and you want a Senior on the build, not a Mid.
The hidden RN line items: Hermes, New Architecture, native modules
Three things every agency post pretends don't exist.
Hermes is now the default JS engine. Since RN 0.70 and standard in every Expo SDK in 2026, Hermes ships smaller bundles and starts up faster than the old JSC engine. You don't pay extra for it, but if you're forking an old codebase that's still on JSC, budget 4 to 8 hours for the migration. Worth it on every metric.
The New Architecture (Fabric + TurboModules) is production-ready in 2026. New apps on RN 0.76+ default to it. If you're starting fresh, you're already on it. If you're migrating an older app, budget 1 to 3 weeks of engineering depending on how many third-party native modules you have and how recently they were maintained. The payoff is real: synchronous native calls, better perf on lists, fewer bridge crashes.
Native modules are mini-projects. Every "we just need a small native module" estimate is wrong by 2x. A single Swift/Kotlin bridge with proper error handling, both-platform parity, and tests is $2,000 to $8,000 of engineering. If your app needs three of them, that's a real line item, not a footnote. This is where the Expo vs bare decision pays off or punishes.
The flip side: Reanimated 3 and Skia have pushed RN's performance ceiling much higher than where it sat in 2022. Apps like Shopify's mobile app, Coinbase, and Discord prove RN can handle complex UIs at 60 to 120fps. The "you have to go native for performance" argument is mostly outdated in 2026, except for very specific use cases (high-frame-rate camera processing, custom video pipelines, AAA games).
How to cut RN costs without painting yourself into a corner
Five rules that compound.
- Default to Expo Managed. Until you have a concrete native-module need Expo can't host, you save money and ship faster.
- Buy commodity infra. Auth, payments, chat, push, analytics. Clerk + Stripe + OneSignal + PostHog gets you 80% of a V1's plumbing for under $200/month and zero engineering time. Same logic that drives SaaS V1 builds.
- Use EAS Update for fixes. OTA pushes ship JS-only fixes in minutes instead of waiting on App Store review. Saves a week of engineering panic per release cycle.
- Pair the engineer with Cursor + Claude Code. Every engineer on Cadence is AI-native by default; the median time-to-first-commit on the platform is 27 hours, and the productivity gap between AI-native and not is real on a JS-heavy codebase like RN. If you're hiring elsewhere, ask in the interview which AI tools they use daily.
- Don't skimp on QA on both platforms. RN bugs hide on the platform you didn't test. iOS-only or Android-only QA passes ship more bugs than they catch. Budget at least 1 week of dual-platform QA before launch.
Worth saying because most posts won't: the cheapest mistake is hiring three different cheap freelancers in sequence. Each one rewrites the previous one's work. You spend $30k and ship nothing. One engineer for 12 weeks who actually finishes is cheaper than three for 4 weeks each who don't.
The fastest path from idea to shipped RN app
Three steps, ordered.
- Spec the V1 in one page. Five core flows max. Defer everything else to V2. If you can't pick five, you don't have a V1.
- Decide Expo vs bare in week one. Default Expo. Only escape if a specific native module forces it. Write down the decision so future-you doesn't relitigate it.
- Book the engineer this week. Not "post a job, screen for two weeks, hire in a month." Book this week. If you don't have an engineer, the fastest path is to book a Mid-tier engineer on Cadence, use the 48-hour free trial to confirm fit, and ship from week one. If you're comparing across mobile platforms, the broader mobile app cost analysis shows where RN sits versus native iOS, native Android, and Flutter.
Try the 48-hour free trial: book a Mid-tier React Native engineer on Cadence, ship for 2 days at no cost, and only pay if you keep them. Weekly billing, replace any week, no notice period.
FAQ
How long does it take to build a React Native app in 2026?
A simple RN app (5-8 screens, auth + Stripe + push) ships in 6 to 10 weeks with one Mid engineer. Mid-complexity apps with chat, maps, or in-app payments land at 12 to 20 weeks. Enterprise apps with custom native modules and 30+ screens run 6+ months and benefit from a Senior or Lead leading the architecture.
Should I use Expo or bare React Native?
Default to Expo Managed in 2026. EAS Build, EAS Submit, and EAS Update cover the build, submission, and OTA-update workflow most apps need. Escape to bare only if you need a native module Expo doesn't have a config plugin for (CallKit, Bluetooth Classic, certain OEM SDKs). The escape hatch from Expo to bare exists if you ever need it; the reverse migration is harder.
Is React Native cheaper than building two native apps?
Yes. Roughly 30 to 45% cheaper than building separate Swift and Kotlin apps for the same scope. Savings come from one JS codebase, one QA pass for most flows, and one engineer instead of two specialists. Native iOS or Android still wins for very specific cases (AAA games, high-frame-rate camera apps, deep platform-specific integrations), but those are the exception in 2026.
What's the cheapest way to ship a React Native MVP?
Expo Managed + Clerk for auth + Supabase for backend + Stripe RN SDK for payments + a Mid-tier engineer for 8 to 12 weeks. Total: roughly $12,000 to $25,000 including SaaS for the first year. If you compare this to the video streaming platform cost stack, the same buy-vs-build logic carries over.
Do I need separate iOS and Android engineers for React Native?
Not for the JS layer; one RN engineer covers both. You do need someone who can debug native crashes on both platforms when they happen, since stack traces in production don't always come from JS. Most experienced RN engineers can read Xcode and Android Studio crash logs. Ask specifically in the interview ("walk me through the last time you debugged an iOS-only crash"); the answer separates the seniors from the mid-fakers.
How much does a Senior React Native engineer cost in 2026?
Full-time in the US, $180k to $230k/year + benefits + equity. Through Toptal or Lemon.io, $100 to $150/hour with monthly contracts. On Cadence, a Senior is $1,500/week with weekly billing and a 48-hour free trial, which works out to roughly half the all-in cost of a US full-time hire when you factor in benefits, equity, and recruiter fees.