If you are a founder with an idea and no technical background, the first question is usually "what will this cost?" The honest answer is a range, and the range depends on a few decisions you control. This post walks through those decisions, then gives you the prices I quote so you have one real number to compare against.
I am a software engineer in New Delhi. I build first versions of products for non-technical founders, so the numbers below come from how I scope and price my own work. Other developers and agencies price differently, and I will not pretend to speak for them.
What an MVP is (and is not)
An MVP is the smallest version of your product that lets a real person do the one thing you promise. It is not a cut-down version of everything. It is a full version of one thing.
If you are building a booking tool for dog groomers, the MVP lets a groomer set their hours and lets a customer book a slot. It does not need reviews, loyalty points, a mobile app and an admin panel. Those come later, once someone is using the first part.
Most budget overruns come from this one point. The price of an MVP rises with every feature you add, so the cheapest way to cut cost is to cut features, not to hunt for a cheaper developer.
The five things that drive the price
Cost follows the amount of work. Five things change the amount of work more than anything else.
- Number of user flows. A flow is one path a person takes, such as "sign up, create a project, invite a teammate". Three flows is a small build. Ten is a different project.
- Accounts and logins. Letting people sign in is a solved problem, but it still adds screens, emails and edge cases like forgotten passwords.
- Payments. Taking money through a provider like Stripe is routine. Splitting money between two parties, as a marketplace does, is much harder and costs more.
- Integrations. Each outside service you connect, such as a calendar, a CRM or an email tool, adds work and adds things that can break.
- AI features. A feature that calls a language model is cheap to demo and slower to make reliable. You need to handle bad answers, slow responses and running costs.
Design matters too. If you already have screens in Figma, the build is faster. If you need the screens designed first, that is separate work and separate cost.
Four ways to build, and what each costs you
You have four realistic options. The money is only part of the comparison.
Do it yourself with a no-code or AI builder. Tools like Bubble, Lovable or Bolt charge a monthly subscription, so the cash cost stays low. You pay in your own time. This works well for a prototype to show people. It gets harder when you need real users, payments and data you cannot afford to lose. I wrote a separate post on choosing between these paths.
Hire a freelancer. A single developer working to a clear scope is often the best value for a first version. The risk is the developer you pick, so it is worth asking the right questions before you hire.
Hire an agency. You get a team, project management and a process. You also pay for all of that. Agency quotes for a first version tend to be higher than a freelancer's for the same scope, partly because of the overhead and partly because of the extra people in the loop.
Find a technical cofounder. This costs little cash and a lot of equity. It only works if you find the right person, and that search can take months.
What I charge
I sell a fixed-price package called the Launch Sprint. You get one working, deployed first version of one core workflow, in three to four weeks.
- Standard Launch Sprint: USD 3,500 to 5,500. This covers up to three core user flows, sign-in, one database, either one payment setup or one outside integration, deployment on your own accounts, and basic analytics.
- With an AI feature or two integrations: USD 6,500 to 9,000. The extra cost pays for the testing and error handling that language models and outside services need.
- Scope & Prototype: USD 500. Before a sprint, we spend time writing a one-page scope with a price and a first milestone. I credit the USD 500 toward the sprint if you go ahead within 30 days.
You also get daily written updates, the code in your own account, a recorded walkthrough, and 14 days of bug fixes after handover. Things I do not take on in a sprint include native mobile apps, two-sided marketplaces with payouts to sellers, and products with heavy compliance rules such as health data. Design from scratch is an add-on.
These are my starting prices. If your scope sits at the edge of a tier, I will tell you before you pay anything.
Where the money goes wrong
A few patterns waste budget again and again.
A vague scope. "An Uber for X" is not a scope. A one-page document listing the flows, the screens and the things left out is. Insist on it before any code is written, from me or anyone else.
Adding features mid-build. Every new idea feels small. A written rule for changes, where each addition gets a price before work starts, protects both sides.
Paying hourly with no cap. Hourly work suits ongoing support. For a first version with a clear scope, a fixed price tells you the worst case up front.
Skipping the basics to save money. Backups, error tracking and a proper domain setup cost little at the start and a lot to add after launch.
How to get a number for your idea
You can get a rough scope in about five minutes. My MVP scope estimator asks nine questions about your product and returns a scope band and a short brief you can copy and send to any developer. It needs no account and no email address.
When you have a brief, send it to two or three developers and compare the answers. Look at the differences between the quotes. If one developer lists things the others missed, you learn something. If one quote is far below the rest, ask what it leaves out.
FAQ
Is USD 3,500 enough for a real product? For one core workflow with a handful of screens, yes. It buys a working first version, not a finished company. Scope is what makes the difference.
Why do some quotes come in at ten times the price? Larger teams, more features, more process, native apps or compliance work. Ask for a line-by-line breakdown and compare scope, not the total.
Do I pay anything after launch? Hosting and outside services bill you directly on your own accounts. Those costs are usually small at the start and grow with usage. I include 14 days of bug fixes. After that, ongoing work gets its own agreement.
Can I start smaller than a full sprint? Yes. The Scope & Prototype step gives you a written scope and a first piece of working software for USD 500, credited if you continue.
What if I have an app from Lovable or Bolt that I need to finish? That is a different job with a different price. The app readiness check shows what to fix first.
Next step
Run the MVP scope estimator, copy the brief, and then either send it to me through the hire page or take it to another developer. Either way you will walk into the conversation knowing what you are asking for.