SHORT ANSWER
For a UK founder, a focused commercial AI MVP with one core journey typically sits around £12,000–£25,000 in the pricing model used on this site. A prototype or narrow validation build can be roughly £3,000–£8,000. Multiple user roles, RAG or agentic workflows, several integrations, sensitive-data requirements or bespoke design can push the build above £25,000. Model and infrastructure running costs should be measured separately per successful customer task.
Answered by Andrew Crossley, AI Product Consultant & Product Partner · Updated 2026-08-22
AI products have variable usage cost as well as build cost. A product can acquire customers and still have weak economics if the model, retrieval and retry path are not measured against what each customer pays.
The phrase 'AI app' covers very different scopes. A one-journey hosted-model MVP should not be priced or planned as if it were a multi-role regulated production platform.
A prototype proves the journey. A focused MVP handles real users, data and a conversion path. More roles, integrations, compliance and AI autonomy make it a larger product rather than an invisible addition to the same budget.
Measure model input and output, retrieval, retries, tool calls and any human review needed for one successful task. Use real pilot inputs rather than a generic per-call assumption.
Multiply the measured successful-task cost by actual task frequency, then add the non-AI infrastructure and support cost. That is the number your pricing and gross-margin assumptions need to survive.
Smaller models, caching, prompt compression, retrieval changes and model routing should be tested against known quality thresholds so a cheaper stack does not silently make the product worse.
FROM EXPERIENCE
Take fifty to a hundred representative tasks from the intended workflow. For each one, record the model and infrastructure cost, whether it succeeded, whether it needed a retry or human correction, and how long the whole task took.
Then divide the total cost by successful tasks and model how many of those tasks an active customer is likely to run each month. That creates a defensible pricing input without pretending one generic token-price figure describes every AI application.
A focused one-journey commercial AI MVP can usually be planned around the £12,000–£25,000 band used on this site. A prototype can cost less; multiple roles, harder AI, complex integrations or specialist compliance can cost more.
Usually test hosted models, prompting, retrieval and workflow design first. Consider fine-tuning when the task is stable, there is enough representative data and an evaluation set can prove the change improves the outcome enough to justify the additional operational complexity.
Measure cost per successful task on representative pilot inputs, multiply by real or conservative task frequency per active customer, then add retries, retrieval, non-AI infrastructure and any human review. Re-run the model whenever prompts, models or customer behaviour change materially.
IN SHORT
Free estimate of what your MVP will cost and how long it will take.
Read moreAI-accelerated MVP delivery, scoped to one journey that earns.
Read moreDone-for-you six-week build turning a founder idea into a launched AI app.
Read moreTurning LLMs, RAG and agents into products that hold up in production.
Read moreA validated, instrumented first product in front of real users.
AI app development in six weeksTHE FRAMEWORK
MORE ANSWERS
No, but it is removing a large part of what product managers used to spend their week doing. AI already writes specs, summarises research, drafts tickets, analyses feedback and builds prototypes. What it cannot do is decide what a company should refuse to build, hold accountability for a metric, or persuade a room. Roles weighted towards documentation are shrinking; roles weighted towards judgement are becoming more valuable.
Startups use AI in three places: inside the product as a feature that completes a user task, inside operations to remove repeated manual work, and inside go-to-market for research, content and support triage. For a small company, internal operational work is often the easiest first use to measure because the baseline task already exists and the team can compare time, quality and cost before and after.
Founders need four categories, not forty tools: a frontier chat assistant for thinking and drafting, an AI-assisted build tool for prototypes and MVPs, a research and synthesis tool for customer and market work, and automation for repeated operational tasks. Pick one per category, use it deeply, and add another tool only when a specific recurring task justifies the subscription and context-switch cost.
Choose an AI product consultant when the uncertainty is the customer problem, scope, AI use case, evaluation, pricing or what should be built at all. Choose a developer when those decisions are validated and the bottleneck is implementation capacity. Early founders often need a product-led build that combines both: senior product judgement to cut the scope and enough engineering execution to ship the one journey that matters.