Free tool · for founders
App development cost calculator
Tick what's in your version one. You get a dollar range and a week range, with every assumption and each item's contribution shown, so you can see what's driving the number and what to cut.
Your estimate
$15,000–$19,500
6–8 weeks with one senior engineer
Rough estimate, not a quote
| Item | Weeks | Cost |
|---|---|---|
| Base: scoping, core workflow, cloud, launch, handover | 3.75 | $9,375 |
| First platform (web) | incl. | $0 |
| Sign-up and login | 0.5 | $1,250 |
| Payments (hosted checkout) | 0.75 | $1,875 |
| Admin dashboard | 1 | $2,500 |
| Total (low end) | 6 | $15,000 |
Assumptions: $2,500 per week of senior-engineer time (my MVP anchor of $15K for about 6 weeks). Low end = the sum above; high end = +30%. Dollars rounded to $500, weeks to whole weeks. Excludes third-party fees, regulated-data work, migrations and custom brand design.
Get a real quote for this scope WhatsApp me
Opens an email with your ticks filled in. You'll get a straight answer on scope, and a fixed quote after scoping.
Typical ranges
What the calculator gives for six common shapes of app, using the weights below. One senior engineer; a second engineer can shorten the calendar but doesn't make the work smaller.
| App | What's ticked | Estimate | Weeks |
|---|---|---|---|
| Lean web MVP | Web, 1 user type, login, payments, admin | $15,000–$19,500 | 6–8 |
| Mobile app on both stores | iOS + Android, 1 user type, login, payments, notifications, admin | $19,000–$24,500 | 8–10 |
| Booking app | Web, 1 user type, login, payments, booking, notifications, admin | $20,000–$26,000 | 8–10 |
| Web MVP with an AI chatbot | Lean web MVP plus a RAG chatbot | $22,500–$29,500 | 9–12 |
| Two-sided marketplace, web only | Web, 2 user types, login, payments, booking, notifications, admin, search | $27,000–$35,000 | 11–14 |
| Large 3-app marketplace | Web + iOS + Android, 3 user types, all ten features | $48,000–$62,500 | 19–25 |
Every assumption, in one table
These weights are my rough estimates for a typical build, in weeks of senior-engineer time, priced at $2,500 a week. They're calibrated to my public anchors: MVPs from $15,000 in roughly 6–10 weeks, and AI features from $8,000.
| Item | Weeks | What it covers |
|---|---|---|
| Base (always included) | 3.75 | Scoping and design doc 0.5, core workflow 2.0, cloud setup and monitoring 0.5, testing, launch and handover 0.75 |
| First platform | incl. | Whichever you build first |
| Any mobile app | +0.5 | Store submission, device testing, store listing |
| Both iOS and Android | +0.5 | One cross-platform codebase, so the second store is cheap |
| Web plus mobile | +1.5 | A second front end on the same backend |
| Each extra type of user | +2.0 | Its own sign-up, screens, permissions and edge cases |
| Sign-up and login | +0.5 | Standard auth provider, roles, password reset |
| Payments | +0.75 | Hosted checkout, receipts, webhooks |
| Booking or scheduling | +1.5 | Availability, time zones, cancellations, reminders |
| Maps or location | +1.5 | Addresses, distances, service areas, permissions |
| Chat | +1.5 | Real-time messages, unread state, moderation basics |
| Notifications | +0.5 | Email and push for key events |
| Admin dashboard | +1.0 | See users, fix data, handle support |
| File uploads | +0.5 | Upload, storage, limits, access control |
| Search and filters | +0.75 | On your main records |
| Analytics | +0.5 | Product events and a basic dashboard |
| Each outside integration | +0.75 | Another system's auth, limits, errors, test accounts |
| AI chatbot (RAG) | +3.0 | Retrieval, prompts, guardrails, an evaluation set from your data |
| AI voice agent | +4.0 | Real-time audio, turn-taking, latency work, transcripts |
| Document reading or scoring (OCR) | +3.0 | Extraction, rubric scoring, human override |
From weeks to a range: the low end is the sum times $2,500; the high end is 30% more. The range skews upwards because forgotten items are always additions. Construx's Cone of Uncertainty white paper puts well-defined projects at about ±25%; estimates from an idea alone can be off by 4x. The full reasoning is in the method article.
Not included: regulated data (health, finance, children), migrating an existing app or data, custom brand or design-system work, and third-party fees (app stores, payment processing, hosting, AI usage). Running costs are covered in what an app costs to maintain each month.
How to use the estimate
- Untick anything the first ten customers could live without. Watch the range move.
- Write the survivors down in a short PRD (template here).
- Send the same document to every builder, and ask each for a breakdown by item.
- Once the scope is written, prefer a fixed price (why, and when not to).
Questions founders ask
How accurate is this app cost calculator?
It's a rough estimate for a typical build by one senior engineer, not a quote. The weights are my own estimates, and the range assumes the scope is written down. Estimates made from an idea alone can be off by much more; a fixed quote needs a written scope first.
Why is the range skewed upwards?
Software estimates rarely come in under, because the items people forget are always additions. The low end is the sum of what you ticked; the high end adds 30%, in line with the roughly ±25% that well-defined requirements and screens can achieve.
Why does a second type of user cost so much?
Each type of user needs its own sign-up, screens, permissions and edge cases, and the two sides have to work together. That's why marketplaces cost more than single-sided apps, and why cutting a user type from version one saves more than cutting any single feature.
What does the calculator leave out?
Regulated data such as health or finance, migrating an existing app, a custom design system or brand work, and third-party fees like app stores, hosting, payment processing and AI usage. Those vary too much by project to tick in a box.