MVP vs Prototype in 2026: What's the Difference and Which Do I Need?
A prototype is a clickable mock-up that tests an idea. An MVP is a working product that tests a market. A prototype costs £0 to £5,000 (roughly $0 to $6,300) and ships in days. An MVP costs £7,500 to £50,000 ($9,500 to $63,000) and ships in 2 to 12 weeks. You build a prototype to find out if your design works. You build an MVP to find out if your business works. Three questions decide which one you need: what are you trying to learn, what evidence do you need, and what is your next step? Most founders skip the prototype stage and pay for it later.
What's the difference between an MVP and a prototype?
A prototype is a clickable mock-up. It looks like a product. It does not work like one. Click a button and the next screen appears, but the data is fake, the logic is fake, the save button does not save anything. The point is to make a design idea visual so you and your users can react to it.
An MVP is a real, working product. Users can sign up. Their data is real. The buttons do what they look like they do. It is on real servers. The product is the smallest version of itself that real users can use end to end.
A proof of concept is a third thing again. A technical spike that proves a hard problem is solvable. Internal only. Often code without a UI.
Side by side:
The pillar covers each in more depth. For this piece, focus on the first two. Prototype and MVP are the two real choices for an early-stage founder.
The three-question test: which one do you actually need?
Most founders pick the wrong one because they skip the diagnosis. Three questions, in order.
Question 1: What am I trying to learn?
If the answer is "do users like this design," build a prototype. If the answer is "will users pay for this," build an MVP. If the answer is "is this technically possible," build a proof of concept first, then a prototype, then an MVP.
Question 2: What evidence do I need?
Opinions from a small sample? Prototype. Behavioural data from real users? MVP. Revenue or strong paid intent (signed up, gave a card)? MVP. A demo for an investor that does not need to actually work? Prototype.
Question 3: What's my next step after this?
If the next step is "iterate on the design," prototype. If it is "show to investors and raise," depends on the round. For pre-seed, a polished prototype plus interview data can be enough. For seed and beyond, an MVP with usage data is the bar. If the next step is "ship to real customers," MVP.
If you can answer all three and they point in the same direction, build that. If they point in different directions, you do not have a clear hypothesis yet. Go back to discovery before you spend money.
When you should build a prototype first
A prototype is the right call in three specific situations.
You're not sure if the user experience will work. You have drawn out the flow on paper. You think it makes sense. But you have not watched a real user try to navigate it. A clickable Figma prototype lets a test user move through your intended flow and tells you immediately where they get stuck. Usually before you have spent a penny on code.
You're not sure if the visual or interaction design lands. You have a brand direction, a colour palette, an idea for the interface. You need to know if it works with the audience before you commit. A static design file is not enough. A clickable Figma version with realistic content will tell you in 20 minutes what static mockups cannot tell you in three weeks.
You need to show investors something visual without committing to build. A prototype is the fastest path from idea to "look, this is what we'd ship." It is not for proving the business works. It is for getting people excited enough to back you while you go and build the real thing.
Tools and rough costs. Figma and Sketch for clickable design mock-ups (free to £12 a month). AI design tools like Google Stitch (free in beta) and Claude Design (included with Claude Pro from $20 a month) generate prototypes from natural language prompts. AI app builders like Lovable, Bolt and v0 all have free tiers, with paid plans from $20 a month. A senior freelancer to build a hand-coded prototype runs £1,000 to £5,000.
Time. A clickable Figma prototype takes 1 to 5 days. An AI-tool prototype takes 1 to 7 days. A hand-coded prototype takes 1 to 3 weeks.
What you learn. Whether the design works. Whether the flow makes sense. Whether the audience responds. You do not learn whether the business works. That is an MVP question.
When you should build an MVP
You are ready for an MVP when one of three things is true.
The hypothesis is "will users pay for this?" Opinions are not validation. "I'd use this" does not mean anyone will pay. The only way to know is to put a real product in front of real users and see if money changes hands. That is an MVP, not a prototype.
You've already validated the prototype and need to test in production. The prototype said the flow works and the design lands. Now you need to know if it survives contact with real users at any scale. That is the next step in the chain. Prototype validates the idea. MVP validates the market.
You need usage data for a fundraise. Investors do not fund clickable Figma files. They fund evidence the business is real. A working MVP with 50 active users and three weeks of usage data is what gets the next cheque.
Costs and timelines, by tier.
Launch Sprint, from £7,500 ($9,500), 2 weeks. A single-product app with simple payment processing built in. Validates a single hypothesis with real, paying users.
Focused MVP, £15,000 to £30,000 ($19,000 to $38,000), 4–6 weeks. A one-sided marketplace or product with messages, reviews, notifications, and a checkout flow. Room for one AI feature.
Complex MVP, £30,000 to £50,000 ($38,000 to $63,000), 8–12 weeks. A two-sided marketplace with multiple AI features. The foundation of a real business.
For the full pricing breakdown, see our 2026 cost guide. For how we run an MVP build, see how our Launch Sprint works.
The four mistakes founders make picking between them
Velocity's view: most founders we work with would have saved time and money by spending two weeks on a prototype before committing to an MVP. The biggest budget mistakes happen when that step gets skipped.
We see the same four mistakes repeatedly. The first one is the most expensive.
Mistake 1: Building an MVP when a prototype would have surfaced the problem.
A founder we worked with spent around £30,000 on an MVP that grew well beyond its original scope during the build. Every "while we're here" decision added a screen, a flow, a feature. The product shipped, working. Then real user feedback came back: too complicated to navigate. Users lost interest. They dropped off.
The complexity was the issue. Not the code, not the design polish. The navigation logic. A clickable Figma prototype of the full intended scope, tested on 10 target users for a week, would have surfaced exactly that feedback. Cost: maybe £500.
The lesson is uncomfortable. It is not usually about building the wrong thing. It is about not catching the right thing soon enough.
Mistake 2: Building a prototype when you needed an MVP.
A founder spends six weeks polishing a Figma file. Shows it to 50 people. All of them say "I'd use this!" The founder concludes the bet is real. It is not. People will say they would use almost anything. Opinions are not behaviour. Until users are signing up, logging in, paying, the validation is theoretical. A prototype is a fast way to know if a design works. It is not a way to know if a business works.
Mistake 3: Calling a prototype an MVP and shipping it.
The Lovable, Bolt, or v0 trap. The founder builds 80% of the product in an AI tool, the demo looks great, and they ship it to real users. The login system breaks. The app forgets what users were doing. The data layer falls over after a few hundred records. Security holes appear. We have audited a number of these MVPs. It is not that the tools are bad. It is that they were never designed to be the foundation of a real business.
Mistake 4: Spending three months polishing a prototype.
Prototypes are throwaway by design. The point is to learn fast and discard. A founder who spends £8,000 and 12 weeks on a prototype has misunderstood the tool. They have built a slow, expensive MVP that does not actually work. If a prototype is taking more than two weeks or costing more than £5,000, stop. You are building an MVP. Either commit to that properly or scale the prototype back.
How much does a prototype cost vs an MVP?
The headline difference is about a factor of 10. Prototype: £0 to £5,000. MVP: £7,500 to £50,000.
Pricing accurate as of June 2026.
A few things to know about the AI tools specifically. Lovable and Bolt use credit systems, which means heavy usage can push monthly costs well above the base subscription. v0 generates frontend code only, no backend or database. None of them produce code you would want to ship to scale. They are brilliant for prototyping. They are not built for production. For the full pricing breakdown on MVPs, see the 2026 pricing guide.
Common questions about MVPs and prototypes
Is an MVP the same as a prototype?
No. A prototype is a clickable mock-up. An MVP is a working product. The most expensive mistake we see is founders treating these as interchangeable. A prototype tests a design idea with internal users. An MVP tests a market with real, paying users. They cost an order of magnitude apart and answer completely different questions.
Should I build a prototype before an MVP?
Usually, yes. Especially if you are a non-technical founder, the UX or design feels uncertain, or you have not watched a real user move through your flow. A few days of prototype work for £500 can save you weeks of MVP rework for £20,000. The exception is when you already have strong validation from prior research and the only remaining question is "will users pay?" In that case, skip to MVP.
Can a Figma file be an MVP?
No. A Figma file can be a beautiful prototype. It can never be an MVP. The defining feature of an MVP is that real users can actually use it, end to end, in production. A Figma file is a picture of a product. It is not a product.
Is a Lovable app a prototype or an MVP?
It depends on what you do with it. As a tool to ship a working demo in a few days for testing or investor calls, a Lovable app is an excellent prototype. As a foundation for a real business with paying customers and growing usage, it is not an MVP. We have audited a number of Lovable apps that hit a wall when real users arrived. The tools were not built for production.
How long should a prototype take?
One to seven days for a clickable Figma prototype. One to seven days for an AI-tool prototype. One to three weeks for a hand-coded prototype with a freelancer. If your prototype is taking longer than three weeks, you have stopped prototyping and started building an MVP slowly. Stop and ask which one you actually want.
How much should I spend on a prototype?
For most founders, the right spend is £0 to £1,500. Free Figma or Google Stitch plus your own time, or a Claude Design or AI app builder subscription for a month. Beyond £5,000, you are paying for something that should be an MVP. The point of a prototype is fast, cheap learning, not polished output.
Do investors want to see a prototype or an MVP?
It depends on the round. For pre-seed, a strong prototype plus customer interview data is often enough. For seed and beyond, the bar is usually an MVP with real users and usage data. The further you go, the less investors care about how it looks and the more they care about whether anyone is actually using it.
What's the difference between a prototype and a proof of concept?
A prototype tests the user experience. A proof of concept tests the technology. A prototype is usually a clickable mock-up shown to non-technical users. A proof of concept is usually internal code that proves a hard technical problem is solvable. Most founders do not need a PoC. If you are not sure whether your idea is technically possible, you do.
Ready to pick one?
If you are working out which to build next, the move is usually one of three.
- Read the founder's guide. Our full guide to building an MVP in 2026 covers the 5 phases, the 4 traps, and the decision criteria in more depth.
- Work out what an MVP would cost you specifically. The 2026 pricing guide breaks down the three tiers, what is included, and where founders overpay.
- Already know you need an MVP and want to scope it. Book a free discovery call and we will scope it honestly. If it is not a fit, we will tell you and point you somewhere it is.
When you are ready.
Discover more informative articles
How Much Does an MVP Cost in 2026? A Founder's Guide to Real Pricing
What an MVP actually costs in 2026, what drives the price, and what no studio will tell you about hidden costs.
.png)
How to Build an MVP in 2026: A Founder's Guide
The 2026 founder's guide to building an MVP. What it is, what it costs, the five phases, and the four traps founders fall into.